Phosphate Glasses

Home , Phosphate glass, Sodium hexametaphosphate

Phosphate Glasses

Francisco Muñoz1, Jean Rocherullé2, Ifty Ahmed3, Lili Hu4

1Institute of Ceramics and Glass (CSIC), Madrid, Spain

2University of Rennes I, Rennes, France

3University of Nottingham, UK

4Shanghai Institute of Optics and Fine Mechanics (CAS), Shanghai, China

Abstract

This chapter is dedicated to the studies on phosphate glasses, from their fundamental

aspects to their most relevant applications of today. P2O5-based glasses have experienced a

continuously increasing number of published works in the last decades and still they

possess a bright potential. Their sometimes intricate structure has made its study a quite

relevant field for the Glass Science community which attracts more and more researchers.

And, on the other hand, the associated difficulties in their preparation on a large scale have

led to the development of specific methods, such as those used for the melting of Nd-laser

glasses. They are particularly known to have a low chemical durability, though the progress

in the optimization of their composition demonstrates that can be very competitive and, in

this respect, we will also pay attention to the improvement of their properties as a result of

their nitridation. The structure and main physico-chemical properties of phosphate glasses

will be reviewed, highlighting the most relevant and well-known applications existing

nowadays, such as sealing and laser glasses, biomedical, as solid electrolytes or for the

storage of wastes.

1.1 Introduction

1.1.1 Phosphorus and Glass Formation

1.1.2 Research and Uses of Phosphate Glasses

1.2 The Structure of Phosphate Glasses

1.2.1 Vibrational Spectroscopies

1.2.2 Nuclear Magnetic Resonance

1.2.3 X-Ray and Neutron Diffraction Techniques

Computational Modelling

1.3 Properties and Applications

1.3.1 Chemical Properties

Oxynitride Phosphate Glasses

1.3.2 Thermal Properties

Low-Temperature Sealing Applications

1.3.3 Optical Properties

Neodymium Phosphate Laser Glass

1.3.4 Biomedical Applications

1.3.5 Electrical Properties

Solid electrolytes for battery applications

1.3.6 Phosphate Glasses for Waste Storage

The study of phosphate melts and glasses dates back to the time when Otto Schott (1851-

1935) was devoted to the investigation of salt melts. Among them, he observed that

phosphate salts formed homogeneous melts as chlorides, fluorides, sulfates or carbonates

do [1]. Then, from 1881, after he started working together with Ernst Abbe in Jena on the

production of new glasses and the study of their optical properties, Abbe wrote to Schott

“The versatility of phosphoric acid is fabulous” [1]. However, due to the technical

impediments in melting large numbers of batches at the time, they soon realized of the

difficulties to attain big, homogenous and crystals-free phosphate glasses. Furthermore, in

trying to produce mixtures of SiO2, B2O3 and P2O5 Schott stated that phosphorus oxide

exhibited a hostile behavior causing opalescence melts. Nowadays, everyone knows, and

has to be aware, of the troubles associated to the melting of phosphate glasses, not to

mention their sometimes extremely high dissolution rate in aqueous media.

1.1.1 Phosphorus and Glass Formation

Phosphorus has an unquestionable role in both Inorganic Chemistry and Biochemistry

thanks to the great number of different bonding configurations that it can adopt, similar to S

[2]. The ground state electronic configuration of P is [Ne]3s23p3 and it is predominantly

present in oxidation states III and V. Phosphorus itself exists in many allotropic forms and,

when combined with oxygen, may give rise to some six oxide structures among which

P4O10, the so-called phosphorus pentoxide is the most important. P4O10 is as so defined

because it forms molecular structures of four PO4 units in a ring arrangement, where three

of the oxygens on each phosphorous are bonded to neighboring P atoms, the fourth being

doubly bonded to phosphorus. A similar arrangement is found in glassy P4O10, described as

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 a three-dimensional network of linked PO4 tetrahedra [2]. Figure 1.1 shows a drawing of a

3D network of fully connected PO4 tetrahedra of P2O5 glass.

O O O P P O P O O O O O O

O O P P O O O O

Figure 1.1: Cartoon of a network of P2O5 glass based on PO4 tetrahedra that are connected

to neighboring groups through P-O-P bridges.

Meanwhile, compounds of phosphorus V are stable against oxidation state +3, through the

formation of a P=O bond, polymeric forms of P4O10 hydrolyze easily to form H3PO4, and

from here derives the extremely high dissolution tendency of phosphate glasses in water.

According to the rule of Goldschmidt [3], by which a ratio of the ionic radii of cation and

anion between 0.2 and 0.4 allows for glass formation, P2O5 is among the oxides that fulfils

this condition, and thanks to its four-fold coordination to oxygens P was classified as a

network-former element by Zachariasen [4], having the highest value of ionic field strength

(2.1) as from Dietzel [5]. The higher difference between the field strength of phosphorus

with those of silicon and boron make the formation of glass from mixtures of P2O5 with

SiO2 as well as B2O3 particularly difficult. However, when modifier oxides are added to

P2O5, a quite broad glass forming range can be attained in most combinations; in MgO-

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 P2O5 system, for instance, the modifier oxide amount may reach 60 mol % while in ZnO-

P2O5 and BeO-P2O5 can be as high as 64.8 and 67.9 mol %, respectively [1]. After addition

of a modifier oxide to the network of P2O5, P-O-P bonds break up and form negatively

charged single bonded oxygens (P-O-), whose charges are compensated with the modifier

cations, and Thilo classified phosphate compositions in ultraphosphates (O/P ratio<3),

polyphosphates (O/P>3) and metaphosphate for the case with O/P=3 [6].

When in solution, phosphate ions tend to form stable long chain structures, and being

highly charged anions they will strongly associate with cations through covalent bonds,

which is thought to be due to the resonance of the -bond in the PO4 tetrahedra [7].

Similarly, it is very common in phosphate glasses to consider the double bonded oxygen as

equivalent to the rest of non-bridging oxygens of the PO4 tetrahedra and in fact they

become undistinguishable as in X-ray Photoelectron Spectroscopy data [8]. It is also

thought that the resonance of the -bond among all P-O bonds stabilizes the structure, most

particularly for the metaphosphate compositions. As it will be seen in detail below in the

section devoted to the atomic structure of the glasses, it is known following Hoppe [9], that

modifier oxide additions to vitreous P2O5 help stabilize the network through formation of

Me-O-P bonds that counteracts the de-polymerization caused by the decreasing number of

P-O-P bonds.

1.1.2 Research and Uses of Phosphate Glasses

The most practical known use of phosphate glasses is the one water softener during

 washing glassware in the form of sodium hexametaphosphate (NaPO3) or Calgon that

helps avoiding deposits of calcium. However, phosphate glasses have remained one of the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 type of glasses that have based their development on fundamental research. According to

Scopus database, while the number of original publications dealing with phosphate glasses

was below 10 per year from 1950 until the beginning of the 70’s, it has increased

exponentially over the last forty years. Due to their much lower chemical resistance than

ordinary silicate glasses, phosphates have met with reticence when looking for new

application fields. However, they also possess some unique properties that allow them to be

employed in fields where no other glass type can achieve the same performance. It is worth

mentioning that there have been two fields of application where phosphates have made

extraordinary achievements, as bioglasses or glass-ceramics [10,11] and laser host materials

[12]. Phosphate based glasses can be formulated to have a chemical composition similar to

that of mineral bone [13] and their biocompatible and bioresorbable properties make them

extremely promising candidates for biomedical applications [14]. They have mainly been

investigated as bioresorbable implant materials for targeted tissue repair applications, with

the main advantages of having easily controllable degradation profiles coupled with ion

release rates and cytocompatibility [13, 15, 16]. These tunable properties are usually

controlled by varying their compositions but can also be controlled to some extent via

surface area modifications [17].

Regarding their application as laser hosts, they have the advantage of having a high

capacity to dissolve rare-earth elements, and thanks to their more open structure may allow

the incorporation of relatively big amounts without significant clustering effects. They also

have large emission cross-section and low non-linear refractive indices, which are ideal for

their application as solid state matrices for the emission of laser radiation [18, 19].

However, the use of glasses as laser hosts requires the production of generally large

dimensions with a very high optical homogeneity and high quantum efficiencies. Even so,

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 neodymium containing phosphate glasses have been successfully applied for the

production of high energy laser radiation in several projects, such as the National Ignition

Facility (NIF) at the Lawrence Livermore Laboratory, in USA, the Gekko-XII in Osaka,

Japan, and the Shenguang projects in China.

Another issue that has experienced a noticeable interest in the last years is related to the use

of femtosecond laser irradiation to produce high refractive index waveguides [20,21] or

even the three-dimensional patterning of metallic nanoparticles [22]. However, the most

well-known, and useful, property that differentiates phosphate glasses is their low to very

low melting temperatures, which makes the melting of phosphates less energy and time

consuming. Nonetheless, the use of elevated temperatures must at the same time limited

due to the ease of volatilization of phosphorus that may cause large losses from the melts as

well as from alkali oxides [23]. On the other hand, their preparation often requires the use

of crucible materials other than platinum-based, at least in a first stage if several melting

steps are needed, due to the corrosion that phosphate glass batches usually cause. In

parallel, it is found that the glass transition temperature is usually in the range of 300-

500ºC, and can even fall below 200ºC in certain cases if only alkalis are used as modifiers

and the glasses are formulated with fluorides. Furthermore, the coefficient of thermal

expansion can also be very high (10-2010-6 K-1), thus resulting in the less novel but,

perhaps, the most studied field of application of phosphate glasses is for the sealing of low

temperature components, e.g. tin and zinc-bearing phosphates [24]. In fact, the main

objective in the search for new sealing glass compositions has been the improvement of the

chemical durability of the glasses while maintaining adequate thermal characteristics for

their use as sealants and, in this respect, the family of oxynitride glasses had its major

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 progress as a consequence of the tremendous increase in the chemical durability of the

glasses that can be achieved by substituting only a part of the oxygen of the glass by

nitrogen [25]. Together with the adequate match of the thermal expansion of the sealant and

those of the components to be sealed as well as a high chemical resistance, it is often

required that the glasses had a very high electrical resistivity and so the formulation of

phosphate-based compositions requires particular modifiers other than alkali oxides that, at

the same time, may lower the viscosity of the undercooled liquid and do not produce an

increase in the softening temperatures, which might become quite challenging as well.

On the contrary, the study of phosphate glasses with high, or relatively high, electrical

conductivity has also been very prolific. In particular, lithium-bearing phosphate based

glasses have been much studied in the last years for their potential applications as solid

electrolytes for rechargeable batteries, due to the numerous advantages they may provide

with respect to the use of dissolved salts in organic solvents [26]. Other studies on

conducting phosphates have dealt with glasses showing protonic conductivity that can be

used as solid electrolytes in intermediate temperature fuel cells [27,28] or because when

containing transition metal elements such as Fe, may present mixed ionic-electronic

conduction and work as electrode materials in rechargeable batteries [29]. Lithium

phosphate glasses generally have higher conductivities than their crystalline counterparts

and can be compatible with most of the electrode materials in use. In particular, lithium or

sodium containing phosphates have been researched as they might be used as solid

electrolytes in secondary batteries and, as it will be seen below, the highest importance of

lithium phosphates as solid electrolytes derives from their relationships with the

development of LiPON electrolytes for lithium micro-batteries [30]. Furthermore, it has

been common in the recent years to study phosphate-based glass-ceramics in which the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 crystallizing phases are related to the NASICON (Na SuperIonic CONductor) structural

type, which account with the highest electrical conductivity among all solid electrolytes to

date [31].

The other characteristic that has make phosphate glasses attractive as an alternative to

silicate and borosilicates is the degree of flexibility of their networks that allows for the

solubility of very high amounts of heavy metal oxides without appearance of phase

separation or devitrification phenomena that could deteriorate their properties. Furthermore,

when those do also contain high amounts of iron oxides the chemical durability increases

up to values that can be even better than silicates [32].

Due to the thoughtful fundamental investigation that has always been carried out previous

to any applicative development, all properties of phosphate glasses have always been

closely related to their atomic structure. As in other glass systems, this has been addressed

by all different existing techniques but there is no doubt that NMR has been a particularly

powerful method to get into the local, or short-range, and the medium-range order structure

of phosphate glasses, facilitated by the high relative sensitivity of the 31P nucleus as

compared to 29Si.

In this chapter, we will make a review of the main structural features of phosphate glasses,

as studied through vibrational spectroscopies, nuclear magnetic resonance as well as X-ray

and neutrons diffraction, and survey all major physico-chemical glass properties paying

special attention to their applications, such as solid electrolytes, laser or sealing materials,

glasses in biomedicine or their use for wastes storage.

1.2 The Structure of Phosphate Glasses

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 It is known since Zachariasen work that the network of phosphate glasses is built up of PO4

units, where at least one of the oxygens forms a non-bridging point with neighboring

tetrahedra that leads to important differences with the structure and properties of silicate or

borosilicate glasses. Then, Van Wazer was one of the first in establishing a tentative

description of the short- and medium-range order structure of phosphates based on the

polymerization degree of their building units depending on the ratio between the

concentrations of modifier and phosphorus oxide [33]. In any case, the interpretation of the

structure in phosphate glasses has always followed the same nomenclature as the one used

in silicates and introduced by Lippmaa [34], which takes into account the number of

bridging oxygens per tetrahedron and gives rise to the well-known Qn terminology. Thus,

the structure of phosphate glasses may be constituted of different arrangements of the

tetrahedra shown in Figure 1.2.

Q3 Q2 Q1 Q0 O O O O

— O — P O — — O — P O — — O — P O- -O — P O-

O O- O- O-

Figure 1.2: Qn structural types in phosphate glasses.

3 Vitreous P2O5 is formed by Q groups where the P=O bond establish disruption points thus

making a weaker network, which at the same time are highly hydrophilic, this being the

reason for the high retention of water in the form of P-OH bonds of phosphorus pentoxide

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 as well as ultraphosphate glasses. When introducing modifier oxides, the P-O-P bonds

break up and form P-NBO bonds where terminal oxygens are linked to modifier cations and

once the metaphosphate composition, for which the O/P ratio equals 3 (MPO3

(M=modifier), is reached only chains or rings of Q2-type groups remain. Upon further

additions of a modifier, the Q2 convert to Q1 and depolymerize the structure until

3- orthophosphate (PO4 ) species appear that, due to their high crystallization tendency with

cations, originate the limit of glass formation in the polyphosphate region.

The structure of phosphate glasses has been tackled by all different analytical techniques,

though Raman and FTIR spectroscopies have been those mostly used. These two

complementary techniques have always allowed for an easy identification of the main

anionic species found in the glasses and how they are affected by the different modifier

cations nearby. Later on, with the advent of the superconducting magnets, the use of NMR

became more widespread and it can honestly be said that solid state NMR of 31P nuclei now

constitutes one of the most common ways of approaching the structure of phosphate

glasses, which, at the same time, is a truly quantitative technique. On the other hand, X-ray

and neutrons diffraction methods have proven to be very valuable for the study of the

coordination numbers and distances at both the short- and medium-range orders.

In the attempt to clarify the structure of phosphate glasses at both the short and medium

range orders, perhaps the studies of Hoppe concerning his specific model for phosphate

glasses have been of greatest importance for the scientific community [9] and, from a

general perspective, the most complete and cited review on the structure of phosphate

glasses may be the one by Brow [35], which accounts for a full set of experimental

techniques and explores all important factors that affect the structure of the glasses.

Raman and FTIR spectroscopies allow for distinguishing the characteristic vibrational

modes of the phosphorus atoms in different configurations with bridging and non-bridging

oxygens, typically vibrations involving P-O-P bonds of neighboring tetrahedra and the ones

- corresponding to the O-P-O bonds in the different species, metaphosphate, (PO3) ,

4- 3- pyrophosphate, (P2O7) , and orthophosphate groups (PO4) . Furthermore, the mode of

P=O bonds in ultraphosphate glasses is also active in both FTIR and Raman spectra. As an

example, in Figure 1.3 the FTIR spectra of ZnO and BaO containing metaphosphate glasses

are shown.

Figure 1.3: FTIR spectra of metaphosphate glasses of the system xZnO-(50-x)BaO-50P2O5

(x=10,20,30,40 and 50 mol %, from top to down, respectively). Reprinted from reference

[36], Copyright (2005), with permission from Elsevier.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 The main peaks of infrared absorption observed correspond to the P=O bonds ca. 1300 cm-

1, the P-NBO bonds at 1000 and 11150 cm-1 and the active modes of the vibrations

involving P-O-P bonds at 950-850 cm-1 and 790-690 cm-1, for the asymmetric and

symmetric modes, respectively [36].

FTIR spectra generally suffer of poor resolution due to the broadness of the peaks attributed

to each active mode, which unfortunately results in very small differences when analyzing

compositional variations. However, Raman spectroscopy can be used to elucidate

frequency changes and variations with composition in a more refined way. Furthermore, the

Raman spectra generally show narrower and more intense peaks in phosphate glasses,

where one can easily observe subtle differences for small changes in composition. As it can

be seen in Figure 1.4, the Raman spectra of a series of metaphosphate glasses of

monovalent and bivalent modifier cations clearly show peaks attributed to the symmetric

stretching mode vibration of the bonds involving bridging oxygens (P-O-P) at ca. 700 cm-1

Raman shift, and those of the phosphorus with two non-bridging oxygens (O-P-O) in Q2-

-1 type PO4 tetrahedra of the metaphosphate polymeric structure at ca. 1200 cm .

Figure 1.4: Raman spectra of metaphosphate glasses 50M2O/M’O-50P2O5 (M=Li,Na;

M’=Ba,Sr,Ca,Mg,Zn). Reprinted from reference [37], Copyright (2014), with permission

from Elsevier.

At ca. 1000 cm-1 a small but in some cases noticeable peak contribution can also be

observed in the spectra of the metaphosphate glasses, which in this case is attributed to the

stretching vibration of the NBO in Q1 terminal groups or pyrophosphate species, appearing

when there is slight deviation of the metaphosphate compositions.

It is generally accepted that Q2 groups in metaphosphate glasses are in the form of long,

also said infinite, chains. However, it must be recognized that ring configurations of

3- tricyclophosphate units (PO3)3 do also appear and they can be easily identified by Raman

as described in Mamedov [38]. As also shown recently by Muñoz-Senovilla et al. in the

alkali and alkaline-earth metaphosphate series [37], the ratio between the intensity of the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 symmetric stretching mode of P-NBO bonds and the sum of those attributed to the same

mode in both chains and rings provides an approximation of the proportion of rings vs

chains in the network that depends of the type of modifier.

Raman spectroscopy can then be used to follow the structure of phosphate glasses through

the ultraphosphate to polyphosphate regions of composition. The Raman spectra of glasses

with composition xLi2O-(1-x)P2O5 can be seen in Figure 1.5. Vitreous P2O5 is clearly

represented by the P=O (1350 cm-1) bond and the one of the bridging oxygens between

-1 phosphorus, P-O-P (700 cm ). With the addition of Li2O, the network depolymerizes and

peaks associated to O-P-O bonds appear below 1200 cm-1, and the one of P-O-P shifts to

higher Raman shifts.

Figure 1.5: Raman spectra of lithium phosphate glasses xLi2O-(1-x)P2O5 (x=0-0.7 mol %).

Reprinted from reference [35], Copyright (2000), with permission from Elsevier.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 At the metaphosphate composition, the active mode of the non-bridging oxygens in Q2-type

-1 groups (PO2) is represented by a narrow and intense peak at 1190 cm , which upon further

1 additions of modifier vanishes to form peaks of pyrophosphate or terminal Q units (PO3),

-1 -1 1050 cm , and orthophosphate species (PO4) at 950 cm .

1.2.2 Nuclear Magnetic Resonance

The isotope of 31P is 100 % naturally abundant and has a high absolute sensitivity with

respect to 1H of 6.6510-2, which makes NMR spectroscopy under magic angle spinning

(MAS) particularly useful for the study of the structure of phosphate glasses. Furthermore,

the 31P is an I=½ nucleus, giving then rise to broad but isotropic signals that in most cases

allow for an easy quantification of all the phosphorus species in the network with much

better resolution than for silicates. Today, 31P MAS NMR is widely used for the

determination of the structural environment of phosphorus nuclei and to obtain the

polymerization degree of the network. In the work by Kirkpatrick et al. [39], the 31P spectra

of glasses with composition xNa2O-(100-x)P2O5 show the distinct structural arrangements

that phosphorus atoms may take depending on the polymerization degree, as it can be seen

in Figure 1.6.

31 Figure 1.6: P MAS NMR spectra of glasses of the system xNa2O-(100-x)P2O5. Reprinted

from reference [39], Copyright (2005), with permission from Elsevier.

A main resonance at ca. -50 ppm attributed to the Q3-type tetrahedra can be seen in a 5 mol

% Na2O containing glass with the presence of a minor one at -30 ppm that corresponds to

2 Q middle-chain groups. It can be appreciated that with further additions of Na2O, the

isotropic chemical shift of the resonances moves downfield at the same time that the Q3

peak intensity decreases and the intensity of the Q2 one increases. Then, a single resonance

2 of Q -type units should be observed at the metaphosphate composition (50 mol % Na2O).

Above this sodium content, the polymeric structure breaks up and forms Q1 with a new

resonance appearing near 0 ppm. Therefore, knowing the proportions of the structural PO4

groups present in the glass one can calculate the modifier and former contents or vice-versa

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 through well-known equations [35,39]. However, in some cases disproportionation

reactions may occur of the type shown below in equations (1) and (2) [35].

Q2  Q1 + Q3 (1)

Q1  Q0 + Q2 (2)

These reactions are mostly typical of polyphosphate glasses, such as in zinc pyrophosphate

glasses [40].

In phosphate glasses, the chemical shift of the 31P nuclei is very sensitive to the type of

modifier cations that are bonded to non-bridging oxygens. As it has been reported in

several works, the chemical shift of phosphorus decreases linearly with the increase of the

modifier’s cationic potential in binary metaphosphate glasses [37, 39]. Due to the particular

structure of the PO4 tetrahedra, the phosphorus atoms become more shielded as the bond

between modifier and oxygens gets stronger. Moreover, if several modifier cations are

present, it has also been proved that the chemical shift varies in a linear way according to a

homogeneous distribution of all modifier types with the structural building units of the

network, as seen for example in glasses with composition (25-x/2)Li2O-(25-x/2)Na2O-

xPbO-50P2O5 [41].

Another interesting picture that NMR can offer on the structure phosphate glasses is

question of whether PO4 tetrahedra are arranged in the form of chains or rings. However, it

is not possible to say, or quantify, how much of the phosphorus belong to either of the

forms. Broadly speaking, one can derive a glass network composed of a higher or lower

proportion of rings, which can be approximated through the analysis of the chemical shift

anisotropy values of 31P resonance [42]. However, this is an issue that has not received

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 much attention but it might be of the highest interest for the medium-range order structure

determination.

Beyond the speciation of the building units of the phosphate glass network, it is also

possible to study their connectivity through dipolar correlation techniques, such as the

Double-Quantum (DQ) MAS NMR experiments. These techniques work by selectively

introducing dipolar interactions between 31P nuclei and depending on their spatial

proximity signals associated to homo- and hetero-nuclear correlations can be observed in a

2D plot [43]. In a polyphosphate glass composition, for instance, homo-nuclear correlation

between Q2 and Q2 middle-chain groups can be seen, as well as those between Q1 and Q1 of

pyrophosphate molecular species. And in the same plot, hetero-nuclear proximity between

Q2 and Q1 groups can also be detected. In Figure 1.7, the 2D 31P DQ-SQ NMR plot is

shown for a glass with composition of zinc pyrophosphate [44].

Figure 1.7: 31P DQ-SQ MAS NMR spectrum of a zinc pyrophosphate glass. The right-hand

side of the picture shows the 1D projection of the Q2-Q2, Q1-Q2 and Q1-Q1 correlations.

Reprinted from reference [44], Copyright (2015), with permission from Elsevier.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 1 1 4- The Q -Q correlation signals are typical of (P2O7) cluster species in pyrophosphate

compositions, though for the glass shown in Figure 1.7 a small proportion of Q2-Q2 can

also be seen, thought to come from small chains remaining in the network. Furthermore, if

an adequate combination of sequences is used and the proportions of the signals associated

to the appearing correlations are known, DQ NMR experiments can help to determine the

length of the phosphate chains in the glass structure, in the form of diphosphate,

triphosphate and polyphosphate or rings arrangements [45].

A more advanced 2D plot of correlations between phosphorus nuclei in phosphate glasses

is the one developed by Fayon et al. and based on refocused INADEQUATE(Incredible

natural-abundance double-quantum transfer experiment) pulse sequences to study the P-O-

P through-bond connectivity [46,47]. The method allows for the determination of direct

bonds between phosphate species by using the J coupling. Even though, in solids, the J

coupling is much smaller than dipolar interactions, the INADEQUATE sequences may

promote polarization transfer through J coupling while avoiding recoupling of dipolar

interactions [46]. Figure 1.8 represents the through-bond SQ-DQ spectrum of a glass with

2 2 composition 0.59PbO-0.61P2O5 in which autocorrelation signals of Q -Q , at -24 ppm in

the SQ dimension, and Q1-Q1, at -8 ppm, tetrahedra can be clearly identified together with

some amount of Q1-Q2, or Q2-Q1 cross-correlation signals in the off-diagonal, as seen in

reference [47].

Figure 1.8: Through-bond SQ-DQ spectrum of the 0.59PbO-0.61P2O5 glass. Reprinted

from reference [47], Copyright (2004), with permission from Elsevier.

1 1 4- Meanwhile the Q -Q connectivity indicates the presence of P2O7 molecular species as

said above, the fact that Q1-Q2 cross-correlation signals can be observed provides

information on the chain length distribution in the glass network. Nevertheless the most

important information that can be obtained through these spectra is the one related to the

direct bonds between the phosphate species by means of the analysis of the chemical shift

values in both the single and double-quantum dimensions of the spectra.

More recently, a new method called DQ-DRENAR (Double-Quantum Dipolar Recoupling

Effects Nuclear Alignment Reduction) has also been tested for the analysis of the structure

in phosphate glasses by Ren et al. [48]. The method, which is based on the REDOR

(Rotational Echo Double Resonance) technique, can also be used for the study of the

connectivity between phosphorus nuclei and the determination of internuclear distances,

and it can be especially useful in combination with refocused INADEQUATE pulse

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 sequence in the case of glasses that have a second glass former or intermediate oxide, for

which the differentiation of the Qn groups gets quite complicated as they tend to overlap

having similar chemical shifts [49, 50].

It is well known that the chemical durability of phosphate glasses can be greatly improved

with the addition of aluminum, which may behave as either former or modifier due to its

intermediate character, a role that depends on the particular composition of each glass. In

this respect the study of the structural environment of Al3+ cations has always been of

importance for the structure-properties relationships. Furthermore, 27Al is a quite accessible

nucleus with a high relative sensitivity and thanks to its quadrupolar character any changes

occurring in the aluminum polyhedra can be easily studied by NMR. These can be in the

form of 4-, 5- and 6-fold coordinated AlOn species, and the AlO6 ones are the most

abundant when a high amount of modifier is present in the composition [51]. Furthermore,

aluminum polyhedra appear to be bonded to phosphate tetrahedra through P-O-Al bonds,

though the number of different species can be very high and the 31P MAS NMR spectra

does not usually offer good resolution to completely characterize and quantify all of them.

In such a case, new advanced methodologies have been proved to be extremely precise for

the separation of the specific P-O-(AlOn) links, as it has been shown by Van Wüllen et al.

using a combination of MQ (Multiple Quantum)-MAS and heteronuclear correlation

experiments (HETCOR) in the study of K2O-Al2O3-P2O5 glasses [52]. Figure 1.9 shows a

representation of the structural model for the potassium alumino-phosphate glasses with

the basic structural units that have been determined through NMR.

Figure 1.9: a) Structural units of the network in a 50K2O-10Al2O3-40P2O5 glass, and b)

representation of the network arrangements. Reprinted from reference [52] Copyright

(2007), with permission from Elsevier.

It is also well known that the properties of the phosphate glasses may also be improved

through the addition of secondary glass former elements, like boron. For example, the ionic

conductivity in alkali borophosphate glasses directly depends on the proportions of the

three-fold and four-fold coordinated species of boron, the so-called BO3 and BO4 groups,

respectively, being maximum for the highest content of BO4 [53].

Finally, the study of the structure of phosphate glasses through NMR may become even

more challenging if one attempts the analysis of the anionic network through 17O NMR.

Even though oxygen accounts for the majority of the nuclei in the glass composition of

oxide glasses, the very low abundance of the 17O isotope and strong quadrupolar moment

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 force to perform the materials synthesis under 17O enrichment. This has been demonstrated

that it can be done efficiently by two methods, the most extended one consisting in the

17 hydrolysis of PCl5 with O-enriched water to form H3PO4 that later is reacted with the

stoichiometric amount of modifier to give a glass [54]; or as shown by Flambard et al.

through the remelting of a NaPO3 glass under 17O-enriched water vapor [55]. However,

very few studies of 17O NMR have been carried out in phosphate glasses to date, but it has

been possible to perform proper quantification of the BO and NBO proportions as well as

getting very useful information on the P-O-P bond angle from the 17O MQ-MAS spectral

17 analysis of NaPO3 glasses [54]. In more complex systems, the use of O MAS and MQ-

MAS spectroscopy may also allow for the determination of the amount of non-bridging

oxygens bonded to either of the modifier cations in the composition or even the distinction

of clustering of the modifier oxides, as shown in Na2O-Nb2O5-P2O5 glasses [56].

The solid-state NMR approach to the elucidation of the structure of glasses, and phosphate

glasses in particular, is continuously evolving and the experience has shown that the

appropriate combination of pulse sequences and methodologies is leading the

understanding of the structure-properties relationships.

1.2.3 X-Ray and Neutron Diffraction Techniques

The materials studied for their interesting technological or scientific properties are often

complex. Typically they are made of multiple elements and do not present an infinite

periodicity. Nevertheless powder diffraction is an important method for the characterization

of these materials but it requires going beyond the phenomenon of diffraction with the

maximum intensities corresponding to the Bragg equation and to favor a total scattering

approach including both the Bragg and diffuse scattering on an equal basis [57]

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Powder diffraction data are collected first, then explicit corrections are made and the

measured intensity is finally normalized by the incident flux. The coherent and normalized

scattered intensity S(Q) is a function of the magnitude of the scattering vector given by

equation (3):

4 πsin 휃 푄 = (3) 휆0

In practice, the S(Q), determined for total scattering studies, is measured over a wide range

of Q-values, however the coherent intensity dies out with increasing Q. Beyond a Q-value

of about 500 nm-1, there are no more features for S(Q). Using laboratory sources, the

-1 -1 maximum Q attainable is around 80 nm from a Cu Kα tube and 220 nm from a Ag tube.

As a consequence, for a better spatial resolution, synchrotron radiation or spallation neutron

sources with short wavelength epithermal neutrons are required. Total scattering data are

Fourier transform to obtain the pair distribution function (PDF) which is fit to give

structural information.

Diffraction methods make it possible to obtain structural information from a radial

distribution of pair distances in which the areas of the peaks are associated with the number

of atoms in the coordination spheres [58]. It has been well established that the first resolved

peak at 0.155 nm is attributed to the P-O bond with a coordination number, NPO, of four.

This number is in perfect agreement with the firmly-established rule that the basic unit of

the network of phosphate glasses is the PO4 tetrahedra.

Similarly, additional information such as the number of metal-oxygen coordination, NMeO,

can be obtained when structural assumptions are made about the adjacent peaks.

Efforts have been made to measure total scattering with high accuracy to extract more

information from the experimental data. Even though these measurements have basically

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 the same requirements than that of any powder diffraction measurements, they have

additional requirements for high quality data. This includes that data must be measured

over a wide Q-range by combining the diffraction data of neutron and synchrotron

radiations. This allows separating the distance peaks of unlike pairs of atoms [60] as well as

the O-O contribution due to edges of the PO4 units, which is separated from the Me-O

distance peak [61]. Thus, using the anomalous dispersion of the X-ray scattering amplitudes

[62,63] and the isotopic change of the neutron scattering lengths [64] several studies of

glassy phosphate materials have been performed. Extensive studies have been carried out

by Hoppe and coworkers on phosphates doped with various ions including La [65], Ga

[66], Fe [67], Pb [68], Zn [69], and Ti [70]. Hoppe et al. developed a set of rules for

network changes upon the addition of modifying atoms to describe the disruption of P-O-P

bonding units as a function of P2O5 content.

Likewise, the truncation of the Fourier integral results in the broadening of the peaks of the

real-space correlation functions T(r). The use of epithermal neutrons (~1 eV) from

spallation sources [71] and of hard X-ray photons (>40 keV) from synchrotron sources

-1 [72], by increasing the maximum Q attainable (Qmax) at almost 500 nm , have considerably

reduce this broadening effect.

The effect of Qmax on the real-space resolution for vitreous P2O5 is illustrated in the

following figure.

Figure 1.10: Illustration of the Qmax influence in vitreous P2O5 [58]. The upper function

results from XRD data [73] (λ0 = 0.0561 Ag Kα). The three lower functions are obtained

from ND data with varying Qmax [10] (λ > 0.022 nm). Solid lines functions are calculated

using the parameters of the Gaussian functions which model the nearest neighbor peaks of

the vitreous P2O5 structure. Reprinted from reference [58], Copyright (2000), with

permission from Elsevier.

-1 A splitting of the P-O distance peak is clearly observed when Qmax exceeds 400 nm . Two

types of bonding should be considered for phosphorous, one to a terminal oxygen atom

(OT) and another to a bridging oxygen atom (OB) [72]. In addition, from neutron diffraction

-1 data for which Qmax is 390 and 470 nm , the O-O peak displays a shoulder denoting two

different O-O distances. If a P-P distance of 295 pm can be determined from XRD data, the

use of ND data with variable Qmax does not make it possible to get the exact value of the P-

P coordination number. The explanation is the presence of P-O and O-O second neighbor

contributions in the P-P peak range.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 The presence of two first-neighbor peaks of P-O distances is always observed when using

ND of great resolving power and this result may be compared to that of O1s XPS (X-ray

Photoelectron Spectroscopy) experiments. Such spectra always display two peaks and as a

result two O atoms species are present in phosphate glasses. Similarly to the analysis of the

total scattering data, the mole fractions of the OT and OB atoms are related to the

corresponding peak areas. Likewise, the ratio of the two oxygen species constitutes a

measure of a continuous depolymerization phenomenon affecting the phosphate network

when adding modifier cations [8, 74, 75].

In a phosphorous pentoxide glass, only threefold corner-linked PO4 tetrahedra have to be

considered. The addition of a modifier cation results in an increase of the number of OT

atoms which constitutes the coordination polyhedra of this cation. Thus, a polymer-like

structure is obtained for a metaphosphate glass composition [75, 76], the chains and/or

rings of the vitreous network being interconnected [77].

Furthermore, the following figure illustrates the influence of the MIIO oxide content,

expressed as the MO/P2O5 molar ratio, on the positions and magnitudes of the P-OT and P-

OB correlations which can be expressed as the P-NBO (non-bridging oxygen) and P-BO

(bridging oxygen) respectively. Another way to best describe the tetrahedral units is the Qn

notation, where n is the number of BOs per unit.

Figure 1.11: Influence of the modifier oxide to phosphorous pentoxide molar ratio (y) on

-1 the lengths of the P-O bonds (ND, Qmax = 470 nm ). The modifier oxides are ZnO for

y=0.61 , PbO for y=1.0 and 1.56 and a mixture of ZnO and PbO for y=1.94. Reprinted from

reference [58], Copyright (2000), with permission from Elsevier.

In vitreous P2O5, each unit possesses one short P-OT bond and three long P-OB bonds and is

connected to three others via bridging oxygen (BO) atoms. As described above, the MIIO

addition disrupts the network and creates NBOs. For the metaphosphate composition there

is an equal proportion of P-OT and P-OB corresponding to two NBOs and two BOs per unit.

Further increasing the MIIO content results in an increase of the NBOs and for a

pyrophosphate glass composition one obtains three P-NBO and one P-BO. It is also

observed that such an increase leads to a longer P-NBO distance. The network becomes

less connected and the NBOs are less strongly bonded to phosphorous.

Hoppe et al. [58] also report on the effect of various modifying cations on the lengths of the

P-O bonds in the case of metaphosphate glass composition. As illustrated by Figure 1.12,

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 the P-OT and P-OB peaks can be differentiated in the case of KPO3 while the splitting is

significantly reduced in AlP3O9. This can be explained, in part, by the cation field strength

which can shift the local electron density away from the BOs. Nevertheless, this effect is

not fully related to the cationic field strength as illustrated by the lowest split of the PbP2O6

3+ glass when compared to that of the LaP3O9 glass, even though the field strength of La is

greater than that of the Pb2+ cation.

Indeed, the change in the metaphosphate behavior, displayed in Fig. 11.2, is the

consequence of both the differences in the P-O lengths and of the widths of the P-OT and P-

OB peaks. The shift of the electron density, affected by the electric field of the neighboring

cation [78, 79] can explain the change in bond length. Ab initio molecular orbital

calculations performed on a series of alkali metaphosphate clusters [80] have shown the

same effects.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Figure 1.12: Influence of the modifier cation on the lengths of the P-O bonds for various

-1 metaphosphate glasses (ND, Qmax = 470 nm ). Reprinted from reference [58], Copyright

(2000), with permission from Elsevier.

Considering the anionic phosphate skeleton, the doubly-bonded oxygen atoms do not

participate to the stability of the structure but tend to coordinate a modifier cation [9]. XRD

studies performed on divalent cation containing phosphate glasses indicate a change of the

Me-O coordination number [81, 82].

This behavior is consistent with the variation of the packing densities of Mg, Zn, Ca and Ba

phosphate glasses which present minima for different compositions [82]. At very low MeO

content, one should observe the highest NMe-O numbers. Experimentally, values of 10 for

Ba, 8 for Ca, and 6 for Mg and Zn are found. The Me-O coordination number must be

lowered when adding further MeO, the experiment confirm that limits of 8 for Ba, 6 for Ca

and 4 for Mg and Zn are reached. The corresponding compositions can be calculated

considering that for the formation of the Me-O-P bridges, the number of NBO per Me ion

has to be equal to the Me-O coordination number. Similarly, the transition temperatures of

alkali phosphate glasses present a minimum at 20 mol % MeO content [83, 84].

As can be seen, the Me-O coordination number plays a decisive role. Nevertheless, this

number is not as well-defined as that for NPO and the uncertainty in the determination of

NMeO limits the exactitude of the model predictions. The best accuracy is achieved when the

Me-O distances are in the range between the P-O and O-O distances at 1.55 and 2.52 nm,

respectively, and do not interfere with the P-O or O-O peaks. A few examples of the

coordination number and of the interatomic distance are given below in table I.

In this table, the NMeO values have been determined from the peak of the correlation

function in closest distance and do not take into account any asymmetry in the coordination

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 polyhedral. Consequently, these values are not necessarily agreeing with the number of

oxygens in a usual MeOn polyhedron.

Table I: Coordination numbers and distance of some atomic pairs.

Atom coordination Distance Reference pair number (nm)

Al-O 6.0 ± 0.3 1.89 ± 0.03 61

Al-O 7.1 ± 0.8 1.89 ± 0.02 78

La-O 7.1 ± 0.5 2.46 ± 0.02 85

Ca-O 7.0 ± 0.4 2.39 ± 0.03 61

Sr-O 6.0 ± 0.2 2.55 ± 0.03 86

Ba-O 8.0 ± 0.5 2.79 ± 0.04 61

Pb-O 5.0 ± 0.4 2.48 ± 0.01 87

3.8 ± 0.3 3.02 ± 0.01

5.0 ± 0.2 2.47 ± 0.03 86

Na -O 5.0 ± 0.4 2.38 ± 0.03 61

K-O 3.0 ± 0.5 2.63 ± 0.03 79

3.7 ± 0.7 2.89 ± 0.03

The knowledge of the medium range order, in other words the reciprocal order of the

largest structural units is limited to diffraction data related to the first peaks [88-90]. As a

result, the specific changes, in relation to structural features, have been described as

functions of the P2O5 content and/or the nature of the modifier cation [91, 92]. Structural

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 features in the neutron structure factors, S(Q), characteristic of intermediate-range (Q ≤ 3Å-

1) order were identified. The addition of an alkali metal such as Na has an effect on the

intermediate-range structure due to destruction of the PO4 network structure. Around the

metaphosphate glass composition a new peak appears at lower Q than the intermediate-

range order peak, which is found in the S(Q)'s of all alkali metaphosphate glasses (M = Li,

Na, K, Rb and Cs), which may be associated with extended-range order. The length scales

of the extended range order increases with the size of M+. These phenomena can be

+ explained by the effects of oxygen atoms, i.e. PO4 chain-like units, ordering around the M .

II The short- and intermediate-range order of M O–P2O5 (M = Mg, Zn) glasses with

compositions around that of the metaphosphate have been studied by neutron diffraction

[93,94]. If the Mg-O coordination number is 6 and almost independent of MgO content, the

average Zn-O coordination number is 5. Both M-O distances increase slightly with the M-O

content from 50 to 60 mol % and the distributions of the distances broaden with the MO

content from 40 to 50 mol %.

In comparison with the structures of zinc phosphate crystals, the distorted five-coordination

of Zn-O may be regarded as the mixture of ZnO4 tetrahedra and ZnO5 polyhedra. As a

consequence, Zn plays a typical dual role as a network-modifier in the process of the

depolymerization of the PO4 tetrahedral network and it acts as a network-former for

forming the linkage of ZnO4 and PO4 tetrahedra. The presence of the two kinds of network

is revealed by the characterization of the intermediate-range order peak and shoulder at

-1 -1 around Q = 1.6 Å and 1.1 - 1.2 Å , corresponding to ZnO4-PO4 and PO4-PO4 topological

connections, respectively.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Likewise, both the average Zn-O distances of the distorted ZnO5 polyhedron and of the Zn-

O coordination number increase slightly when increasing the ZnO content from 59.3 to

69.7 mol %.

These small changes in the short- and intermediate-range order associated with the

chemical ordering of PO4 and ZnO4 tetrahedra and ZnO5 distorted polyhedra probably

explain the anomalous changes in physical properties such as the increasing Tg behavior at

ZnO concentrations above 60 mol %. As it was mentioned in section 1.2.2, zinc phosphate

glasses can also develop disproportionation reactions at around the pyrophosphate

composition, giving rise to out of the trend changes in the properties that could be

explained through the phosphate groups speciation as well as the change in the coordination

number of zinc.

Computational Modelling

Various methodologies can be used to fit the data from crystalline compounds. The most

familiar is the Rietveld refinement [95] which can be operated to both X-ray and neutron

diffraction data. Nevertheless, fitting diffraction data related to glassy materials is more

questionable and different processes have been conceived. These models are usually

generated by Reverse Monte Carlo (RMC) [96,97], Molecular Dynamics (MD) [98,99] or,

more recently, by Empirical Potential Structure Refinement (EPSR) [100,101].

However, these methods are computationally intensive and the calculations are limited to a

maximum of a few hundred atoms, and by consequence their agreement with experiment

can be poor in the low-q region of the data. In addition, even ab initio techniques still

require some choice in terms of, for example, the density-functional, and this can affect the

results.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Considering the Ca(PO3)2 metaphosphate glass, the short range structure has been

investigated using both X-ray and neutron diffraction and then modelled using the reverse

Monte Carlo method [102]. This approach of combining computer simulations with more

than one set of experimental data avoids to fail in interpreting correctly the experimental

results and provides to obtain more robust and detailed structural information. Based on the

geometry of the basic phosphate unit, a model has been refined to match the experimental

diffraction data. Thus, six interatomic correlations have been differentiated and fitted to

finally obtain two values of the Ca-O bond length, i.e. 2.35 and 2.86Å, and a calcium

coordination number of 6.9. This latter value is coherent with distorted polyhedral units

such as capped octahedral or capped trigonal prisms. In comparison, calcium

metaphosphate crystals present a capped trigonal prism structure. In addition, it appears that

most of the non-bridging oxygen atoms are bonded to two calcium atoms, creating Ca

clusters with a Ca-Ca correlation distance which is a function of the sharing mode between

two adjacent Ca coordination polyhedra, i.e. edge-sharing or corner-sharing. Nevertheless,

there is no quantitative way to describe the Ca clusters in terms of their shape.

It should be noticed that the main advantage of the RMC method is that it is comparatively

easy to use and the program is readily and freely available. In addition, data sets from

different techniques (e.g. diffraction, EXAFS, NMR, etc.) can be ‘fitted’ simultaneously

while respecting quantitatively their different experimental errors. Furthermore it is

relatively easy to build extra constraints into the refined model by, for example, adding

coordination number constraints obtained from, e.g. NMR methods, or requiring the model

to have agreement with local atomic arrangements known from, e.g. EXAFS

measurements. A disadvantage is that it is difficult to build in molecular structure to the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 model apart from using rather crude constraints that risk trapping the simulation in local

minima.

In the past ten years, the application of molecular dynamics simulations in the field of

bioactive glasses has improved both the identification of structural features and the

fundamental understanding of composition-bioactivity relationships in these materials

[103]. If most of the work has focused on biosilicate glasses, several investigations have

been carried out on the effects of incorporating fluorine on the structure, and hence the

bioactivity, of fluorinated phosphate based glasses [104,105]. From these studies, it has

been established that atomistic simulations have revealed details of bulk structural features

which affect the glass dissolution and thus its bioactivity, such as the connectivity network

and the tendency to form chains, rings and clusters. The replacement of an oxygen atom in

a PO4 tetrahedron by a non-bridging fluorine atom causes a slight reduction in the glass

network connectivity, which is likely to increase the bioactivity. A second effect, namely

the segregation of the glass network into modifier-rich and network-rich regions with an

attendant decrease in bioactivity, is likely to be insignificant due to the sizable amount of P-

F bonding observed. As a result, fluorinated phosphate-based glasses do not suffer the same

decrease in bioactivity as fluorinated silicate-based glasses, and can be considered as strong

candidates for biomaterials.

1.3 Properties and Applications

In the following section we will put our main interest in the most important applications

that phosphate glasses have nowadays with relationship to their properties, such as

chemical, thermal and mechanical, electrical or optical. Phosphate glasses can be used for a

variety of applications within very diverse fields, though in most cases a suitable

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 combination of several of their properties is needed and which can be found through the

formulation of complex compositions. Nevertheless, due to the intrinsic importance of the

atomic structure on the interpretation of the variation of properties with composition, most

relevant studies on properties are always based on simple systems.

1.3.1 Chemical Properties

Discussing chemical properties of phosphate glasses, researchers have always focused on

their dissolution behavior in aqueous media. This has been a recurrent topic when studying

the properties of phosphate glasses and remains as the long-standing issue to be resolved

during the design of new compositions for any application. The origin of their ease for

dissolution resides in the highly hygroscopic nature of the phosphate units forming the

glass network, especially to protonated species, and likely related to the high stability of

phosphate compounds into solution [106]. While, in most cases one looks for the highest

chemical durability, there are fields where a strict control of the dissolution rate of the

glasses must be imposed independently of whether high or low, such as in biomedical

applications, resorbable materials, controlled drug release, etc.

Comparatively speaking, phosphate glasses are more soluble than silicates, but the

mechanisms of dissolution are similar in both, and when studying dissolution rates the

same important factors are taken into account: composition, surface area, temperature, pH

and time. Furthermore, the most common way to study the dissolution kinetics has been

through the weight loss of a bulk glass sample against time, normalized to its surface area

and keeping constant conditions of the ratio between the surface of the sample and the

volume of solution [107]. Sometimes, convenient agitation of the media can be used, but

for simplicity static condition are normally employed. On the other hand, while most of the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 studies carry out the tests at room temperature, especially those indicated for biomedical

studies in simulated body fluid solutions, it has also been usual to perform kinetics at

different temperatures even up to 90-100 ºC either in static or dynamic conditions, if

extreme environments need to be reproduced in a short period of time [108]. Other methods

include the standardized dissolution tests that simulate the corrosion of nuclear waste

glasses and are described in the ASTM C-1285-94 specifications, or the so-called Product

Consistency Test (PCT), to follow the dissolution of powdered glasses at 90 ºC.

It is generally observed that the dissolution of phosphate glasses takes place congruently

into two separated stages when studying the weight loss against time: first, a period that

depends on t1/2 and is attributed to the diffusion of water through the glass surface; second,

a constant loss stage depending linearly on t [109]. However, as numerous factors are

involved in the whole process of dissolution (composition, temperature, pH, etc.) any

change in one of them may affect the duration of each of these stages. Bunker et al.

proposed a mechanism by which only until the polymeric phosphate chains are completely

hydrated (stage t1/2) the dissolution of the whole chains may take place (t-dependent stage).

This was proposed as in agreement with the amount of acid that the dissolution of the

phosphate glasses needed to keep a constant solution pH, which at the same time allowed

for a calculation of the length of the dissolved chains, being around 40 units [109]. Other

authors, however, have proposed that after the hydration of the glass surface, with exchange

of protonated species by modifier ions (Li+ or Na+), there must be a breakage of the P-O-P

bonds of the hydrated layer that results in the dissolution of the glass [110]. Similarly,

Döhler et al. stated that the hydrolysis step follows the hydration of phosphate groups and,

furthermore, that the contribution of the hydrolysis of P-O-P bonds in the mechanism

depends on the phosphate content after studying the dissolution of glasses with

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 compositions xP2O5-(100-x)/2CaO-(100-x)/2Na2O (x=35-55 mol %), for which the higher

the amount of P2O5 the higher the extent of hydrolysis is [111]. In their work, Döhler et al.

observed that the solubility of the glasses, as studied from powdered glasses (125-315 m)

in pH=7.4 tris-buffered solutions at 37ºC, decreases with the decrease of the phosphate

content in the glasses, and they suggested that shorter chains would be less hydrolysable.

This situation is completely different to what has been seen in sodium or lithium binary

phosphate systems, where the opposite trend occurs [112, 113], in which a decrease of the

solubility results with the increase of the P2O5 content. Therefore, it is clear that both type

and content of modifiers, as well the ratio between phosphate and modifiers, give rise to a

different behavior and likely variations in the mechanism. What seems to be clear is that,

concerning the nature of the modifiers, the chemical durability of the glasses will increase

with the higher ionic field strength of the cations and that the overall solubility will not

only depend on the hydrolysis step of the P-O-P bonds but also on the strength of the metal

to oxygen bonds.

Delahaye et al. also point out to another influencing factor on the t1/2 first stage of

dissolution related to the ionic strength of the solution [107]. The authors have also

concluded that a diffusion controlled step occurs for short times of dissolution; however,

they argued that this would happen as a result of the increase in the ionic strength of the

solution that limits the rate of dissolution of the glasses until it keeps constant for longer

durations. Furthermore, one should also take into account if an alteration of the surface of

the glass takes place through formation of hydrated layers or precipitation of crystalline

compounds that may be forcing changes in the mechanism of dissolution. For example, the

dissolution behavior of ZnO-P2O5 bulk glasses in distilled water at temperatures of 30 to 90

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 ºC for 72 h was classified according to the precipitated products appeared in solution or

onto the sample surface after the leaching tests, which showed decreased solubility with the

decrease of P2O5 content, and even lower solubility for the highest temperatures [114]. In

the highest phosphate containing glasses the formation of a corrosion layer onto the surface

dramatically decreases the dissolution rate of the glasses.

Oxynitride Phosphate Glasses

If there has been one way by which phosphate glasses have achieved a considerable

increase of their chemical durability it is clearly their nitridation. The nitridation of

phosphate glasses, or the partial substitution of oxygen by nitrogen, was originally develop

by Marchand at the University of Rennes 1 by thermal ammonolysis of a phosphate base

glass under an NH3 flow [115], in an attempt to establish a comparison with the already

known oxynitride silicate liquids (SiAlON) that play an important role in the joining of

silicon nitride ceramics and related phases [116]. The literature on oxynitride phosphate

glasses is not as extensive as that of their silicate counterparts because silicon oxynitride

glasses have always been of great interest not only for glass scientist but also for ceramists.

However, phosphate melts possess the advantage that they can be nitrided at much lower

temperatures and, under adequate conditions, produce glasses with much better

homogeneity and transparency, which may expand their possibilities of application.

Actually, in principle all considered applications for phosphate glasses can also be

considered for oxynitride phosphates, with higher chemical durability.

The first nitrided phosphate glass by Marchand was a NaPO3 composition at 700 ºC for 100

h in order to achieve the maximum allowed nitrogen content in the glass, reaching nearly

10 wt. % that corresponds to a glass with composition NaPO2.01N0.66 [115]. Thus, the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 substitution of nitrogen for oxygen takes place in a stoichiometric way through the

exchange of 2N3- anions by 3O2-, resulting in the following reaction, given by the reaction

below:

NaPO3 + xNH3  NaPO3-3x/2Nx + H2O

The reaction must take place at temperatures higher than 600ºC, above which

decomposition of NH3 may proceed in order to react with the melt; however, it has to be

realized that temperature is another limiting factor due to the fact that this cannot surpass

800ºC because phosphorus can be easily reduced and form colored inhomogeneous glasses.

In any case, to be able to achieve high nitrogen contents in a short period of time the

temperature should be as high as possible, which is related to a low melt viscosity [117].

The first remarkable observation by Marchand was the increase of Tg of the glasses after

nitrogen incorporation, which was of more than 100ºC with respect to that of the sodium

metaphosphate glass. This change of Tg was already explained through the increased cross-

link density and covalent character introduced by the two nitrogen species that may appear

in the oxynitride glasses, the dicoordinated nitrogen, P-N=P (Nd), and the tricoordinated

one, P-N<(P)2 (Nt), as seen later by X-ray Photoelectron Spectroscopy [118].

Soon after the experiments of Marchand, the study of oxynitride phosphate glasses became

an interesting topic of research within the field of phosphates due to the interesting

modifications of both their structure and properties. The first studies of the dissolution of

oxynitride phosphate glasses were performed by Day and coworkers at the University of

Missouri and showed that the dissolution of the oxynitrides was up to 10000 times slower

than that of the oxide parent glasses, as shown by Bunker et al. in glasses of the system

Na2O-BaO-Al2O3-P2O5 [119]. They observed a linear decrease of the dissolution rate at

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 temperatures between 30 and 70ºC with the nitrogen content, that decreases to almost that

of soda-lime glasses for the highest nitrogen contents, and interpreted the results as a

consequence of a much slower diffusion of water throughout the glass surface and a

mechanism involving direct hydrolysis of the P-N bonds. In such a case, the dissolution

rate changes from a mechanism controlled through the square root of reaction time to

another depending directly on t. Furthermore, they concluded that the dissolution of the

oxynitride compositions is less sensitive to pH changes, thus leading to glasses that can

resist better in a wider range of conditions.

Further studies demonstrated the impact of nitrogen on several other properties, such a

decrease of the coefficient of thermal expansion and increases in density, refractive index

and hardness [120-122], and all results can be explained on the basis of the increased cross-

linking of the glass network through the new bonds created. The effect on viscosity was

also studied, showing an increase with the content of nitrogen [121]. Despite the observed

trend, authors determined lower activation energies for the viscous flow in the oxynitride

sodium metaphosphate glasses; however, this issue was not much researched in detail at the

beginning. More recently, Paraschiv et al. have started a systematic study of the

thermodynamics and kinetic fragility study in nitrided phosphates with interesting results

that show a decreasing fragility of the glasses with nitrogen content that might be related to

an inhomogeneous microstructure composed of regions with different rigidities [123].

Another way of incorporating nitrogen into the glasses was attained through the addition of

metal nitrides during the melting of a sodium phosphate glass [124]. The results showed the

same modification of properties of oxynitride glasses though in this case further changes in

composition of the glasses takes place as secondary metals are being added to the glass,

such as Mg or Al, and due to the higher melting temperatures needed to dissolve the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 nitrides inhomogeneities may often result and therefore this method has not been much

explored.

The studies on the structure of oxynitride phosphate glasses through XPS of the O1s and N1s

energy levels allowed for the determination of the reaction mechanisms of ammonolysis.

Both BO and NBO oxygens decreased their amount as the nitrogen content is increased in

the glasses; however, the ratio between them, BO/NBO showed a linear decrease in all

cases though with slight varying rates depending on glass composition [125]. Brow et al.

concluded that the nitrogen for oxygen substitution takes place on both BO and NBO

simultaneously, without preference for a substitution in either of the oxygen types in the

network. At the same time, Marchand et al. defined the two substitution rules that

constitute the basis for the nitridation: 1) 2Nt=3BO and 2) 2Nd=2NBO + BO [118]. As

shown above, Nt and Nd refer to the tri-coordinated and di-coordinated nitrogen atoms,

respectively. From the analysis of the N1s XPS spectra of the Nt and Nd relative proportions,

one may calculate the expected values of the BO/NBO ratio assuming the above rules and

following equation (4):

BO/NBO = (BO/NBO)0 – [1.5Nt/(2 – Nd)] (4)

where (BO/NBO)0 represents the initial BO/NBO ratio in the parent glass [118]. Using the

above rules, Le Sauze et al. showed a good correlation when studying mixed alkali

oxynitride metaphosphate glasses with composition Li0.5Na0.5PO3-3x/2Nx glasses. Figure

1.13 shows the experimental BO/NBO ratio as determined by XPS, black squares, and the

calculated values, dotted line, following Marchand’s rules [126].

Figure 1.13: Calculated vs experimental BO/NBO ratio in Li0.5Na0.5PO3-3x/2Nx glasses.

Reprinted from reference [126], Copyright (2000), with permission from Elsevier.

Furthermore, N1s XPS results generally show a higher amount of Nt bonds at the first stages

of nitridation, i.e. low nitrogen contents, which might be related to an initial preference for

the substitution of the bridging oxygens alone, contrary to that proposed before by Brow et

al. [125]. As the nitrogen content increases, both Nt and Nd proportions become similar and

it may also happen that Nd surpasses the content of Nt. As it was shown by Muñoz et al.

comparing alkali (‘NaPON’), mixed alkali (‘LiNaPON’) and mixed-alkali lead

metaphosphate glasses (‘LiNaPbPON’), the variation of the Nt/Nd ratio does also depend on

the glass composition, showing higher Nt values for the sodium metaphosphate system

[127].

The glass network structure has been also well characterized by 31P NMR, which may also

contribute to a precise determination of the reaction mechanisms. It is known that PO4

groups transform into PO3N and PO2N2 units after nitridation and in their work, Le Sauze

et al. showed the complete evolution of these structural species as a function of the nitrogen

content [126]. Based on their results and the analysis done by 31P DQ MAS NMR, they

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 proposed that the nitridation first proceeds by the substitution of nitrogen for oxygen to

form PO3N groups, homogeneously distributed in the glass network, and then, continues by

the formation of PO2N2 on the previously existing PO3N-PO4 links [126]. Thus, the authors

concluded that the mechanism assumes that oxynitride regions grow at the expense of the

oxide ones, which was further verified in LiNaPbPON glasses by Muñoz et al. [128,129].

Figure 1.14 shows the evolution of the P(O,N)4 tetrahedra as a function of the nitrogen

content in Li0.25Na0.25Pb0.25PO3-3x/2Nx glasses [128].

Figure 1.14: Evolution of the proportions of P(O,N)4 tetrahedra with nitrogen content (x) in

oxynitride glasses with composition Li0.25Na0.25Pb0.25PO3-3x/2Nx. Reprinted from reference

[128], Copyright (2003), with permission from Elsevier.

While the proportion of PO4 groups progressively decreases with x, PO3N and PO2N2

increase with nitridation, though the one of PO3N groups remains higher than that of PO2N2

for all x. These proportions vary similarly in all glass compositions but some differences

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 can be found depending on the type of modifiers, particularly at high nitrogen contents

[127]. Furthermore, the fact that for the highest nitrogen contents, a higher amount of Nd

type bonds is found and indicates that PO2N2 tetrahedra will host a higher amount of

dicoordinated nitrogen species, likely due to a higher stability forced by steric effects. The

above data were also used to determine that the nitridation mechanism is constituted by a

system composed of two pseudo-first-order consecutive reactions as shown by Muñoz in

[130]. First, the reaction with NH3 gives rise to the formation of PO3N from PO4 and then,

PO2N2 form from the substitution of nitrogen for oxygen in the PO3N groups that are linked

to neighboring PO4 with the subsequent formation of new PO3N. Both reaction steps were

determined to have the same activation energy of ca. 150 kJ.mol-1 [130].

Finally, the last recent finding was also achieved by Muñoz et al. through the use of 17O-

enriched NaPON glasses [131]. In these glasses, the structure was studied by 17O high-field

MAS and MQ-MAS NMR and the authors were able to determine that two types of non-

bridging oxygens are located on the P(O,N)4 tetrahedra, while for the NaPO3 reference

glass a single type of NBO is found. After the modeling of the 3Q-MAS spectra taken at

18.8 T, it was possible to determine of the NMR parameters of each site and quantify their

population. The results revealed a second type of NBO whose amount increased with the

nitrogen content in the NaPO3-3x/2Nx studied glasses, being of ca. 80 % of the total NBO

content for the maximum N/P ratio of 0.68. As the new non-bridging oxygen contribution

would be that coming from the NBO located in the PO3N and PO2N2 groups formed during

nitridation, the authors were able to calculate the proportion of those given the fractions of

each of the structural units and assuming that PO3N groups have two NBO and one BO

while PO2N2 have no BO left, obtaining a quite good agreement with the experimental

quantification.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 The most extended studies of oxynitride phosphate glasses have dealt with their application

as sealing glasses, not only for their high chemical durability but also due to their still

adequate softening temperature and thermal expansion for the sealing of low melting point

metals [132]. However, the fact that their dissolution rate can be controlled and adjusted as

needed through the choice of the right composition and nitrogen content, let us think that

oxynitride phosphates have a much bigger potential for present and future applications,

such as in the field of bioresorbable phosphate glasses for medical applications [133]. Last

but not least, it is worth mentioning here the studies on the electrical conductivity of

lithium oxynitride phosphate glasses and their relationship with the LiPON electrolytes that

are used in lithium microbatteries [30]. As it will be seen below, these amorphous thin-film

electrolytes possess the same increase in ionic conductivity and chemical and mechanical

resistance thanks to the partial substitution of oxygen by nitrogen as the lithium oxynitride

phosphate glasses have [134].

1.3.2 Thermal Properties

The most important and studied thermal properties of phosphate glasses are glass transition

temperature and thermal expansion behavior, as they have been of the major interest for

their application as low temperature sealing glasses [135]. Other properties have received

much less attention and have been studied for fundamental interest only, such as the heat

capacity [136,137]. Thermal conductivity has normally been determined when being a key

factor in the behavior of the glass under working conditions, such as in the case of the Nd-

doped phosphate laser glasses [19], where a large thermal conductivity contributes to an

increased thermal shock resistance during laser operation as will be detailed below.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Low glass transition temperatures are usually associated with high thermal expansion

coefficients and have a clear relationship with the two main structural factors, the field

strength of the modifier cations and the polymerization degree of the glass forming

network. The less interconnected network in phosphate glasses resulting from the P=O

bond of the PO4 building units has a tremendous impact on the Tg and simultaneously on

the thermal expansion of the glass. Furthermore, even minor amounts of water can remain

dissolved into the glass and affect Tg with an additional substantial decrease. It is also well-

known in lithium and sodium phosphate glasses that Tg first decreases with the addition of

alkali oxide down to a minimum around 20 mol %, then continuously increases with further

additions of modifier [138]. This was attributed to a re-distribution of the Li-O bonds

depending on the alkali concentration rather than to an abrupt change in the coordination

number of lithium cations that give rise to an increase of Tg through the formation of cross-

links of O-Li-O polyhedra [139], being in agreement with the structural model of Hoppe

[9]. Nevertheless, the glass transition temperature increase is always associated with the

ionic field strength of the modifier cations, though there are exceptions like when using

cations like Zn or Pb. These elements are known to provide relatively high chemical

durability values while maintaining low transition temperatures as well as high coefficients

of thermal expansion. However, as it has been shown recently, a linear relationship appears

between Tg and the product of the cationic potential of the modifier and its coordination

number in metaphosphate glasses, where now the Tg of the zinc metaphosphate

composition does not result abnormally low compared to the values in alkali and alkaline-

earth containing glasses [37].

As indicated above, the glass transition temperature of alkali phosphate glasses can be

lower than 300ºC although the chemical resistance of simple phosphate compositions, like

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 those or even with some additions of alkaline-earth elements, is rather low. Zinc, lead and

also bismuth can be used in combination with alkaline-earth elements to provide low Tg and

high thermal expansion at the same time that relatively good durability values.. However,

these compositional changes do not normally meet the specific requirements for the

application of phosphate glasses and so the most effective option is the formulation of

glasses with Al2O3. This can greatly affect the thermal behavior of the glasses but

undoubtedly will improve their chemical resistance. Furthermore, the incorporation of even

minor amounts of alumina in the glass composition may help to reduce the crystallization

tendency, being also advantageous for the control of the sealing process through

maintaining adequate fluidity. In this sense, it is worth mentioning the important role of the

viscosity of the glass, not only for their production, but also as a determinant factor that

must be strictly controlled when the phosphate glasses needs working as a solder glass

[140].

Low-Temperature Sealing Applications

Glasses are known to be suitable for the production of mechanically reliable and vacuum-

tight seals with metals and ceramics. The direct wettability of materials by glasses and the

viscosity behavior of glasses are among the critical properties. The stability and mechanical

strength of a glass seals must be ensured by limiting the mechanical stress in the glass

component at temperatures met during production and use. In other words, the thermal

contractions of the two sealing components should match each other below the transition

temperature of the glass. In addition, glasses must very often fulfil other requirements such

as possessing high electrical insulation, chemical resistance to the environment or specific

optical properties. Glasses are generally classified according to their composition, such as

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 silicates, borosilicates, alumino-borates, lead and zinc borates, antimonates, vanadates or

phosphates. However, they can be also classified according to their thermal expansion and

temperature characteristics into "hard" and "soft" glasses [141]. Hard glasses have low

thermal expansion coefficients, i.e. α < 510-6 K-1, whereas soft glasses possess higher

thermal expansions, i.e. α > 810-6 K-1. As a general rule, glasses of high expansion possess

relatively low softening and working temperatures, while low-expansion glasses have

higher softening points. Glasses with particularly low softening temperatures are referred to

as "solder" glasses. This is because they are used to join glass to other glasses, ceramics, or

metals without thermally damaging the materials to be joined. Soldering is generally

carried out in the viscosity range of η = 103 - 105 Pas of the solder glass, which

corresponds to various temperature ranges, depending on the chemical compositions.

Among the various glass compositions which have been developed, the Pb-based sealing

frits, such as PbO–B2O3–SiO2 or PbO–ZnO–B2O3 glasses, are the materials of choice for

most commercial sealing operations including, in late years, plasma display panel (PDP) or

vacuum fluorescent displays (VFD). These materials offer the advantage of low viscosities

for which the softening point is <400°C, coefficient of thermal expansion in the range 9-

10×10-6K-1 and flow over an extended period of time without simultaneous crystallization

when fired as a fine powder (10-40 μm) during the sealing process. However, they suffer

from a severe drawback which is the high PbO content (typically more than 70 wt.%), a

component with deleterious health and environmental effects. For these reasons, a search

has been conducted for glass compositions with the lowest possible Tg as alternatives to Pb-

based frits. One such system is the SnO–ZnO–P2O5 (SZP) ternary [20]. Both the aqueous

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 durability and crystallization of the SZP frits may be made comparable to that of Pb-based

systems by adequate design of the composition.

The development of solder glasses with very low soldering temperatures is limited by the

fact that reducing the temperature generally means increasing the coefficient of thermal

expansion. This effect is less pronounced in devitrifying solder glasses and it can be also

avoided by adding inert fillers with low or negative coefficients of thermal expansion (e.g.

ZrSiO4 or β-eucryptite).

Vitreous solder glasses have to be distinguished from devitrifying solder glasses, according

to their behavior during the soldering process. The properties of solder glasses do not

change during soldering; upon reheating the solder joint, the temperature dependence of the

softening is the same as in the preceding soldering process. Contrary to vitreous solder

glasses, devitrifying solder glasses crystallize. They change into a ceramic-like

polycrystalline material during soldering and the viscosity increases by several orders of

magnitude during crystallization so that further flowing is suppressed.

Only limited attention has been devoted to the controlled crystallization of phosphate

glasses to produce glass-ceramic materials. Much of this work has been aimed at the

production of biomedical materials based on calcium phosphate for applications involving

bone replacement and dental implants [141]. The crystallization behavior of a number of

phosphate systems, including Na2O-CaO-P2O5, Na2O-BaO-P2O5, Na2O-A12O3-P2O5 and

Li2O-BaO-P2O5, have been investigated likewise the effect of a number of potential

nucleating species, such as TiO2, ZrO2, Y2O3, La2O3, Ta2O5, WO3 and platinum [142].

Among these additions, only platinum was found to be effective at promoting bulk

crystallization. Relatively high thermal expansion coefficients in the range of 16.2 -

22.510-6 K-1 have been achieved. A devitrification study conducted on calcium phosphate

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 glasses containing a number of oxide and fluoride additions have shown that TiO2, in

conjunction with A12O3, may also be employed to promote bulk crystallization [143].

Thermal expansion data for a number of phosphate glasses and glass-ceramics are given in

Table II, with their corresponding chemical compositions, which have been taken from

references [141] and [144].

Apart from characteristic data, such as coefficient of thermal expansion (CTE), glass

transition temperature or elastic properties, the importance of the design of the seal should

not, however, be underestimated. A wrong design is usually associated with mismatch in

coefficient of thermal expansion, producing tensile stresses in the glass. These usually

manifest themselves as cracks, emanating from the interface into the bulk of the glass,

although if the stresses are not high enough to cause cracking initially, time-dependent

failure may occur due to the influence of static fatigue. A good design must take into

consideration any possible mismatch in thermal expansion and seek to place the interface

under compression [144].

A common feature of all compression seals can be illustrated by glass-to-metal seals which

are used to provide electrical connections to components that must be hermetically sealed.

In this typical application, electrical connectivity is provided through metallic pins arranged

within an opening of a metallic shell that is sealed with a glass to insulate/isolate the pin

from the housing. A single-pin, concentric glass-to-metal seal is depicted in Figure 1.15.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Figure 1.15: Schematic draw of a glass-to-metal compression seal.

The choice of materials is an important factor in reducing the risk of glass fracturing.

Compression seals use an outer shell material with a CTE that is greater than the glass and

the CTE of the pin that closely resembles or is lower than that of the glass. As such, the

glass is subjected to nominal radial compression on cooling that tends to reduce the

possibility of generating high tensile stresses. As a result, the glass body is kept under

overall radial pressure after the sealing. This pre-stressing protects the glass body against

dangerous mechanical loads and guarantees robust seals. Because the compressive stress of

the glass is compensated by a tensile stress in the metallic shell, the wall must be

sufficiently thick (at least 0.5 mm even for small seals) in order to be able to absorb such

tensions permanently. Compression seals can be produced as hard glass or soft glass seals.

If the difference between the thermal expansion of the metallic shell and that of the sealing

-6 -1 glass is significantly higher than 510 K , an additional pre-stressing of the glass body

may result.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Table II: Composition and characteristic temperatures of phosphate sealing glasses (Data selected from references [141] and [144 ]).

Sealing Thermal Temperature Code Composition (mol %) temperature expansion range (°C) (°C) (10-6 K-1) Na2O K2O BaO ZnO B2O3 Al2O3 SiO2 P2O5 PbO Others G1 25.00 25.00 50.00 212 34.7 100-250 G2 35.00 15.00 50.00 221 30.0 100-250 G3 15.00 50.00 35Ag2O 266 26.8 100-250 G4 38.00 3.00 56.00 3 Fe2O3 317 26.0 100-250 G5 38.00 56.00 6 Fe2O3 303 25.6 100-250 G6 34.00 8.00 58.00 313 24.0 100-250 G7 34.00 4.00 58.00 4 Fe2O3 350 22.0 100-250 G8 30.00 10.00 60.00 374 20.0 100-250 G9 50.00 50.00 310 17.1 100-200 G10 58.80 41.20 255 15.3 100-200 G11 20.00 40.00 40.00 345 13.2 100-200 G12 20.00 50.00 30.00 325 12.1 100-200 G13 26.47 9.72 6.63 57.18 450 11.6 20-400 G14 8.11 30.57 61.32 330 11.6 100-200 G15 30.00 50.00 20.00 305 11.0 20-200 G16 15.78 29.74 11.87 42.61 460 9.6 20-400 G17 26.84 32.88 40.28 CuO 540 7.1 20-400 G18 40.42 7.87 5.38 46.33 475 7.0 20-400 G19 2.82 42.48 4.77 8.08 41.85 500 6.4 20-400 GC1 40.00 10.00 50.00 22.5 25-250 GC2 20.00 30.00 50.00 16.2 25-250 40.30 CaO 14.1 20-185 GC3 7.40 7.80 39.10 5.50 TiO2 *G=Glass; *GC=Glass-Ceramic

It is well known that phosphate glasses deviate from the “normal line” behavior that

silicate glasses have in the representation of the relative partial dispersion and the Abbe

number [1]. Due to their higher chemical and mechanical resistance, and the fact that

they follow a predictive value of the partial dispersion vs the Abbe number, silicates are

the standard optical glasses. Phosphate glasses generally possess higher partial

dispersions for the same Abbe number than silicate glasses and, together with borates,

which have a negative contribution, can be used for the correction of the secondary

spectrum of silicate glasses [1]. In the field of optical lenses, phosphate glasses have

found application in Precision Glass Molding, where the use of low Tg glasses helps in

reducing the temperature and time of processing. In any case, the low chemical

durability of phosphates to environmental moisture, again, limits their use and most of

the works today remain searching for new compositions with adequate thermal and

optical properties while having acceptable chemical durability [145]

An advantage of phosphate glasses is, however, their ability to be melted with fluorides

for the production of fluorophosphates glasses. Particularly, high fluorine containing

phosphate glasses are excellent for optical glasses with positive anomalous dispersion

that can be used for the substitution of CaF2 single-crystals in the reduction of

chromatic aberrations [1,146].

Phosphates and fluorophosphates in particular are good candidates as athermal glasses

where near zero coefficient of refractive index (dn/dT) and low thermo-optical constants

can be easily achieved, usually better than in silicate glasses. Furthermore, when glasses

have a negative value of the index variation with temperature, phosphate can be used for

the elimination of wave aberration due to temperature changes [1]. Negative values of

dn/dT can be attained in metaphosphate glasses of modifier cations with low ionic field

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 strength, such as in K, Na, Ba or Pb metaphosphates due to their large thermal

expansion coefficient [147]. In fact, potassium and barium are modifier elements

generally used in the formulation of laser phosphate glasses as it will be seen below

[12].

Other interesting features of phosphate glasses are their non-linear refractive index

values (n2) and fluorescence linewidths. Phosphates have moderate non-linear index and

low linewidths while fluorophosphates glasses have even smaller n2 and linewidths,

which in addition to their higher UV transparency and lower phonon energy makes

them very appropriate for luminescence applications [148]. Nevertheless, phosphates

are very prone to present water in their compositions and very few amounts of hydroxyl

ions produce very drastic non-radiative fluorescence quenching that may limit their use

if the glasses are not properly melted [149].

Neodymium Phosphate Laser Glass

Neodymium doped phosphate laser glass (simplified as Nd:phosphate glass) is widely

used in high power laser facilities. There are two kinds of Nd:phosphate laser glasses:

one is for high peak power laser application, another is for high average power

application. Neodymium ions have eight absorption bands from UV to infrared range

due to the 4f3 electronic energy level transitions. According to the energy diagram, the

3+ 4 Nd ion is used mainly for 1 μm laser application due to the transition between F3/2

4 and I11/2 energy levels. This is a typical four energy level laser which has low laser

threshold and laser efficiency is less sensitive to thermal effects. Phosphate glass is used

as the matrix for Nd3+ ions as it has medium phonon energy, high solubility to rare earth

ions, large damage threshold and superior spectroscopic properties compared with

silicate glass matrix. Therefore, neodymium phosphate glass is a priority laser material

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 of choice for 1 μm laser facilities. In addition, platinum inclusions can be easily

oxidized and dissolved as platinum ions in phosphate glasses, thus preventing the

formation of Pt inclusions, which is a key parameter to prevent the laser damage for

high power laser application [19].

The composition design for Nd:phosphate glass should consider first its application. For

different application purpose, there will be different composition design.

The figure of merit (FOM) of high peak power Nd:phosphate laser glass is given in

equation (5) [150].

() Q   FOM  abs0 em ex laser n 2 (5)

Where Δλabs is the absorption bandwidth, σem is the stimulated emission cross section,

ηex is energy extraction efficiency, n2 is the nonlinear refractive index, Q is

3+ concentration quenching factor and τ0 is fluorescent lifetime at zero Nd ion

concentration.

The most important parameter of high peak power Nd:phosphate laser glass is the peak

stimulated emission cross section σem. It is calculated by J-O method [151,152]. Large

peak stimulated emission cross section is preferred to achieve high gain efficiency.

Another important parameter is nonlinear refractive index n2. To prevent the risk of

nonlinear optical damage, smaller n2 is preferred for high peak power Nd:phosphate

laser glass. The high peak power Nd:phosphate laser glass is usually a kind of

metaphosphate glass (P/O=3) [19]. The peak stimulated emission cross section of

Nd:phosphate laser glass is sensitive to glass composition. The main composition of

Nd:phosphate laser glass is P2O5-Al2O3-M2O-MO [19]. P2O5 content in high peak

power Nd:phosphate glass is about 55-65 mol %. The stimulated emission cross section

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 3+ of Nd ion usually increases with P2O5 content from 45 to 70 mol % P2O5 in phosphate

glasses [19]. M2O can be K2O or a mixture of alkali oxides, while MO can be one

alkaline oxide or a mixture of alkaline-earth oxides. Small cationic field strength such as

those of K+ and Ba2+ ions results in large stimulated emission cross section of Nd3+ ion

in phosphate glass. MgO content is used to achieve lower nonlinear refractive index but

it has larger cationic field strength and it decreases peak stimulated emission cross

3+ section of Nd ion. Al2O3 content is used to improve the chemical durability and to

control the thermal expansion coefficient of Nd:phosphate laser glass in reasonable

range for fabrication. But it decreases the stimulated emission cross section of Nd3+ ion.

In addition, small amounts of high valence oxides such as Nb2O5, La2O3 and Y2O3 are

used in Nd:phosphate laser glass to improve the chemical durability and vitrification

behavior of the glass.

Figure 1.16 presents compositions for both research and commercial Nd:phosphate laser

glasses [19]. It is found that the commercial Nd:phosphate laser glass is usually

metaphosphate glass with O:P ratio of 3.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Figure 1.16: Composition range diagram of Nd:phosphate laser glasses. Reprinted from

reference [19], Copyright (2000), with permission from Elsevier.

The figure of merit of high average power Nd:phosphate laser glass is shown in

equation (6).

KK(1 ) FOM  1c (6) tm E

Where E is Young’s modulus, Klc is the fracture toughness, K is thermal conductivity, v

is Poisson’s ratio and α is linear thermal expansion coefficient.

For high average power laser applications, the Nd:phosphate laser glass works in lower

repetition rate such as 0.1-15Hz. It results in the accumulation of thermal gradients and

may cause distortion of the laser beam and even the failure of the laser host. In such a

case the thermo-mechanical properties of the glass should be optimized. The second

goal is to get high gain coefficient. In order to optimize the thermal mechanical

property, lower thermal expansion coefficient and large thermal conduction are

preferred. The main high average power composition is P2O5-Li2O-Al2O3-SiO2. A large

amount of Al2O3 is contained in this glass in order to get lower thermal expansion

coefficient. Li2O is used to increase the thermal conductivity of glass. As a result, high

average power Nd:phosphate laser glass has a smaller stimulated emission cross section

compared with high peak power Nd:phosphate laser glass.

Another important property of neodymium laser glass is the gain coefficient. It is

expressed in equation (7),

gN ()   loss (7)

where σ(λ) is the emission cross section at laser wavelength λ. For Nd:phosphate glass,

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 the maximum emission cross section in the near infrared region is around 1053 nm. ΔN

is Nd3+ inverted population density. It is determined by the fluorescent lifetime and

fluorescence effective bandwidth of this state. αloss is the attenuation at the laser

wavelength. From eq.(7) it is known that in order to achieve high gain coefficient, large

stimulated emission cross section, long fluorescent lifetime and lower attenuation are

preferred. These three parameters are dependent not only on glass composition but also

on the fabrication process.

The stimulated emission cross section of Nd3+ ion can be evaluated from the

measurement of absorption and emission spectra of Nd3+ ion in glass by method

reported by Krupke [153] and based on Judd-Ofelt (J-O) time [151,152].

64 2e 2 n ( n2 2) 2 AS   (8) JJ ''3hJ (2 1)3 9 JJ

I() d eff  (9) I()p

4  AJJ '  2 (10) 8cn  eff

Equations (8) to (10) are used to calculated the stimulated emission cross section of

3+ Nd ion. In equation (8), AJJ’ is probability for a transition from initial J manifold to

terminal J’ manifold, n is refractive index, λ is wavelength, h is Plank constant, J is the

total angular momentum of initial level, J’ is the total angular momentum of terminal

level and SJJ’ is line strength of an electric-dipole transition between initial J manifold to

terminal J’ manifold. In equation (9), Δλeff is effective linewidth of fluorescent

spectrum, I(λp) is fluorescent intensity at peak position, I(λ) is the fluorescent intensity

at wavelength λ. And finally, σ is the stimulated emission cross section with c as the

speed of light, in equation (10). The radiative transition probability from J state to J’

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 state, AJJ’, can be obtained from absorption spectrum with equation (8). The effective

linewidth, Δλeff, can be calculated by equation (9) from emission spectrum. Stimulated

emission cross section of Nd3+ ion can be calculated from equation (10). It is known that

large AJJ’ and small Δλeff will result in a large stimulated emission cross section.

Figure 1.16 indicates the effect of cationic field strength, P2O5 and Al2O3 contents on

the spectroscopic parameters of Nd3+ ion in phosphate glasses. It is clearly shown that

the contents of Al2O3, alkali oxide and alkaline oxide can affect the stimulated emission

cross section effectively. For alkali oxide and alkaline oxide, the smaller the cation field

3+ strength, the larger the stimulated emission cross section of Nd ion. Al2O3 content will

decrease the stimulated emission cross section and increase the effective linewidth of

emission. Radiative decay rate (that is radiative transition probability) changes in the

same trend as the stimulated emission cross section with glass composition. The

emission bandwidth increases with the increase of cation field strength from K+ to Li+

and from Ba2+ to Mg2+ in Fig. 3. There is in generally increase trend of stimulated

emission cross section with the increase of P2O5 content from 45 mol % to 70 mol % in

Figure 1.17.

Figure 1.17: The effect of cationic field strength, P2O5 and Al2O3 contents on the

spectroscopic parameters of Nd3+ ion in phosphate glasses. Reprinted from reference

[19], Copyright (2000), with permission from Elsevier.

The fluorescent lifetime is related to radiative transition rate (Arad) and non-radiative

transition rate (Wnr). It is determined by glass composition, fabrication process, and the

3+ Nd ion concentration. In equation (11) the non-radiative transition rate Wnr is closely

related to the fabrication process. The larger the non-radiative transition rate, the shorter

the fluorescent lifetime. Radiative transition rate Arad is determined by composition of

glass. Large Arad corresponds to short fluorescent lifetime. Radiative lifetime is the

inverse of radiative transition rate in equation (12), which is determined by the glass

composition. The fluorescent lifetime of Nd3+ ions is inversely related to Nd3+ ion

concentration as express in eq. (13), where N is the Nd3+ ion concentration, Q is

concentration quenching factor, which is physically equivalent to the Nd3+ ion

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 concentration needed to reduce the fluorescent lifetime to one half of its zero

concentration limit.

  1 rad WA nr)( (11)

1  rad  (12) Arad

   0 (13) 1 (NQ / )2

From equation (13) it is known that the fluorescent lifetime decreases with the increase

of Nd3+ ion concentration. As the fabrication process of Nd:phosphate glass is highly

moisture sensitive. Dehydroxylation process is necessary to remove hydroxyl groups in

glass melt and to achieve long fluorescent lifetime of Nd3+ ion in phosphate glass matrix

[154].

The attenuation at the lasing wavelength (1053 nm in Nd:phosphate glass) is determined

by the impurity contamination, absorption of Nd3+ ion in infrared range, and scattering

caused by bubbles or inclusions in glass. The ppm level of transition metal ions (such as

Cu2+, Fe2+) and rare earth ion impurities (such as Dy3+,Pr3+, Sm3+ ) can be harmful to the

attenuation at lasing wavelength. In order to get the attenuation as low as possible, the

impurities in raw materials and impurity contamination during fabrication process

should be controlled. The extinction coefficient of Cu2+ ion in Nd:phosphate laser glass

is as large as 2.7×10-3 cm-1 [155]. Cu2+ ion is most harmful transition metal ion to

attenuation of Nd:phosphate laser glass of all impurities. Its content in glass should be

controlled in ppb level.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 The contribution of Nd3+ ion itself to attenuation can be expressed in equation (14). At

room temperature for 4.2×1020 cm-3 of Nd3+ ions concentration, absorption coefficient

of Nd3+ ion at 1053 nm is at the level of 7×10-4 cm-1 [155].

2576  (T ) 1.03 1020 [Nd 3 ]exp( ) (14) Nd T

-1 3+ Where Nd (T) is the loss in cm , [Nd ] the neodymium concentration and T the

temperature in K. Typical application of high peak power Nd:phosphate glass is for

laser inertial confinement fusion facility, such as National Ignition Facility (NIF) in US,

Laser Megajoule (LMJ) in France and Shen Guang facility in China. These facilities use

commercial high peak power Nd:phosphate glasses. They are LHG-8 from Hoya

Company, LG-770 from Schott Company, N21 and N31 from SIOM in China as well as

KGSS-0180 from GOI in Russia. Table III lists the main parameters of 4 kinds of high

peak power Nd:phosphate laser glasses. They are used in NIF in US, LMJ in France,

Shen Guang facilities in China and high power laser facility in Russia, respectively.

Table IV shows the high power laser facilities worldwide and the high peak power

Nd:phosphate glasses used inside. More than 3000 pieces of large size Nd:phosphate

laser glasses have been used in the largest laser facility-NIF in US.

Table III: Main parameters of high peak power neodymium phosphate laser glasses

from Hoya [19], Schott [152], Russia (GOI) [156] and SIOM [157].

 /10-20cm2 3.4 3.8 3.6 3.9 3.6

* rad /sec 348 365 351

eff /nm 26.5 25.5 26.5 25.4

*d /g/cm3 3.40 2.87 2.83 2.59

* nd 1.5758 1.540 1.5296 1.5067 2.83

n1053nm 1.5652 1.533 1.5201 1.4991

Abbe number 65.2 65.8 66.5 68.4

-13 n2 /10 esu 1.30 1.18 1.12 1.01

Tg /°C 500 450 485 460 1.1

 /10-7/K(20-100°C) 110 115 115 116 460

dn/dT /10-7/K -42 -43 -53 -47 116

dS/dT /10-7/K 19 14 6 11 -40

k /W/m.K 0.55 0.56 0.58 0.57

E /GPa 56.4 50.0 47

*parameters that vary with Nd2O3 concentration.

TableIV: Nd:phosphate glasses used in high power laser facilities

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Laser facility Nd:glass Number of laser beam Volume of glass

Omega-EP in US LHG-8 60 2.3L

Gekko-ＸII in Japan LHG-8 12 2.7L

NIF in US LHG-8, LG-770 192 15L

Shen Guang II N21, N31 8+1 3-7L

Shen Guang III prototype N31 8 7.6L

Shen Guang II upgrade N31 8 12L

Shen Guang III N31 48 15L

LMJ in France LHG-8, LG-770 240 15L

Table V gives 5 kinds of high average power Nd:phosphate laser galsses. HAP-4 is

from Hoya Company. APG-1 and APG-2 are from Schott Company. NAP-2 and NAP-4

are from SIOM. These glasses have relative smaller stimulated emission cross section

but with lower thermal expansion coefficient and larger thermal conductivity compared

with those of high peak power Nd:phosphate laser glasses in Table II.

The typical application of high average power Nd:phosphate laser glass is laser

processing, such as laser peening [12]. In recent years, with the development of petawatt

(1015 watt) high power laser, the Nd:phosphate glass laser is used as pumping source in

the petawatt laser system. There is an increasing demand on high average power

Nd:phosphate laser glass.

Table V: Main parameters of high average power neodymium phosphate laser glasses

parameters HAP-4 APG-1 APG-2 NAP-2 NAP-4

 /10-20cm2 3.6 3.4 2.4 3.7 3.2

* rad /sec 350 385 464 380 400

eff /nm 27.0 27.8 31.5 27.0 29.0

*d /g/cm3 2.70 2.64 2.56 2.76 2.60

* nd 1.5433 1.5370 1.5127 1.542 1.530

n1053nm 1.5331 1.5260 1.5032 1.536 1.523

Abbe number 64.6 67.7 66.9 67 66

-13 n2 /10 esu 1.21 1.13 1.06 1.22 1.10

Tg /°C 486 450 549 478 545

 /10-7/K(20-100°C) 72 99.6 64 96 71

dn/dT /10-7/K 18 12 34 -8.7 19

dS/dT /10-7/K 57 52 76 36 50

k /W/m.K 1.02 0.78 0.84 0.76 0.86

E /GPa 70 71.0 64.0 58 67

1.3.4 Biomedical Applications

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Due to the chemical instability of phosphate based glasses, various modifier oxides have

been explored in order to make phosphate glasses compatible with applications in

biomedicine, including metal oxides such as Na2O, CaO, MgO, Fe2O3, Sr2O3, Al2O3,

CuO and TiO2 [158]. Among them, Fe2O3 and TiO2 have been reported to have greater

resistance to hydration due to their network strengthening capability via their cross-

linking effect [159,160]. As such, these metal oxides have been explored quite widely in

order to try and impart control over the glass dissolution rates [161]. Phosphate glasses

have also been fabricated into continuous fibers (Phosphate Glass Fibers or PGFs) via

both a melt and pre-form drawing process (see Figure 1.18). The melt drawing

technique usually involves melting glass frits in a crucible containing bushing tips with

small holes at the exit [17]. Molten glass is then allowed to flow gravitationally through

the bushing tips and the fibers formed are then collected on a rotating drum in order to

produce continuous fibers. The pre-form fiber drawing process is a two stage process

which involves production of a glass pre-form rod followed by heating the pre-form rod

in a furnace to above its Tg. At this temperature a molten gob forms at the end of the

pre-form which falls away from the tip due to gravity producing a fibrous strand which

is then collected onto a rotating drum [162].

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Figure 1.18: SEM image of melt drawn quaternary PGFs (fibers with composition of

50P2O5-40CaO-16Na2O-5Fe2O3 in mol %). Reprinted from reference [17], with

permission from “Hot Topics in Biomaterials” as agreed by Future Medicine Ltd.

Successful production of continuous PGFs depends on several parameters such as melt

temperature, melt viscosity, fragility index, oxygen/phosphorus ratio and drawing speed

[163]. Proper selection of glass composition is essential so that the bonds present within

the glass melt are of sufficient strength to withstand the stresses applied during the fiber

drawing process. Additionally, further heat treatment (also known as annealing) of the

fibers produced can be employed in order to relieve internal stresses present within the

fibers, thus achieving a more stable configuration [164]. The process parameters of fiber

production have also been found to play an influential role in varying the fiber

mechanical and dissolution properties. For example, highly fragile glasses are

manufactured into fiber via a melt drawing process where rapid cooling of the molten

strand is essential. On the other hand comparatively less fragile glasses can be processed

into continuous fiber via both the melt draw and solid pre-from fiber manufacturing

routes [163].

PGFs have some advantageous features over their bulk form, such as excellent

mechanical properties and higher surface area. These properties can also be varied by

controlling the fiber dimensions via process variables such as the diameter of the

bushing exit holes, melt temperature (hence viscosity of the glass), mass flow and

drawing speed [163]. The dissolution rates for PGFs have been seen to increase with

decreasing fiber diameter, due to the vast increase in surface area. For example, Ahmed

et al. [15,165] reported that degradation rates of phosphate glass fibers (within the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 ternary P2O5–CaO–Na2O systems) increased significantly compared to bulk glass of the

same composition. Furthermore, additions of metal ions (such as Na2O, CaO, MgO or

Fe2O3 etc.) within the glass structure enabled further control over their dissolution rates.

For example, addition of Fe2O3 content (from 1 to 5 mol %) within the iron phosphate

glass system (P2O5–CaO–Na2O–Fe2O3) exhibited a significant reduction in the glass

dissolution rates from ~0.00045 to ~0.00004 mg.cm-2.h-1 [165].

More recently, core/clad resorbable PGFs with varying metal ions (such as Ti2+ and

Fe3+) were successfully manufactured by Ahmed et al. [166]. This type of core/clad

fiber was earlier limited to optical glass fibers only. The main advantage associated with

core/clad fiber was the additional control conferred over their ion release profiles. It was

also shown that hollow fibers could be produced if a comparatively faster degrading

composition of the inner core glass was selected. The core/clad glass fiber production

process involved the following steps: a) manufacture of glass billets with varying

compositions (preferably with similar thermal expansion profiles), b) co-extrusion of

the stacked glass billets by placing the cladding glass underneath the core glass (see

Figure 1.19a), and c) the core/clad fibers (see Figure 1.19b) were then drawn from the

core/clad pre-form via the previously mentioned pre-form drawing technique.

Figure 1.19: a) Cross-sections of an extruded core/clad preform (the values indicate the

cut distance of the preform into discs) and b) an extruded core/clad preform and

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 resultant fibers. Reprinted from reference [166], Copyright (2015), with permission

from De Gruyter.

Mechanical properties of the glass fibers produced via any of the above mentioned

techniques depend on the molecular orientation of the glass components along with the

bonds created within the chemical compositions [167,168]. For example, the bulk glass

consists of chemical bonds which are usually in isotropic form; whereas during glass

fiber manufacture the network bonds are converted into anisotropic form as the PO4

tetrahedra align in the direction of the pull [169,170].

PGFs have also been utilized to reinforce bioresorbable polymers to fabricate

bioresorbable composites [171,172]. As the mechanical properties of PGFs can be

varied by simply changing their diameter as well as their chemical compositions, a wide

range of PGF-reinforced composites can be produced with tailored mechanical

properties to match both cortical and cancellous bone [173,174]. For example, very

recently PGFs with high mechanical properties (for example, tensile modulus ~ 70 GPa

and tensile strength ~1.2 GPa) were produced from boron doped PGFs (with the

formulation 45P2O5-16CaO-5Na2O-24MgO-10B2O3) [175].

Furthermore, the orientation of PGFs within the composites can also have an influential

role on their mechanical properties. For example, continuous unidirectional (UD) fiber

reinforced composites produced superior mechanical properties than randomly (RM)

oriented fiber reinforced composites (properties achieved from tensile modulus for RM

and UD composites ranged between ~6.7 and ~9 GPa, respectively) [176]. Other factors

that also influenced the overall mechanical properties of PGF reinforced composites

were fiber volume fraction, fiber length, distribution and strength of the fiber/matrix

interface [173,174].

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Phosphate based glasses in the form of powder, bulk, rod, and fiber containing varying

modifier ions such as iron, zinc and copper have been explored for a range of

biomedical and tissue engineering applications [165,177,178]. These metal ions have

been found to be beneficial for promoting various biomedical functions. For example,

iron ions are well known to promote cell attachment and differentiation as well as to

participate in redox reactions for certain types of proteins (such as, cytochrome,

myoglobin, etc.) [165,177].

Similarly, Zinc ions were found to stimulate protein synthesis in osteoblast cells and

also increase ATPase and ALP activity [179] and copper ions have been reported to

have antibacterial properties, stimulate angiogenesis and proliferation of human

endothelial cells [180]. As such, Zn doped PGFs in the form of 3D-scaffolds were

investigated for the construction of muscle organoid units [177] and Cu-doped PGFs

were found to prevent bacterial colonization and reduce the number of viable bacteria in

the local environment [178].

Researchers also investigated iron doped PGFs as potential cell delivery vehicles for

cell transplantation purposes and utilized their morphology to orientate muscle

precursor cells along the axes of the fibers to form myotubes (see Figure 1.20) [165]. 4-

5 mol % Fe2O3 was reported to be most favorable for cell attachment and differentiation

due to the enhanced chemical durability of the PGFs and controlled ion release profile

[165].

Figure 1.20: a) Attached muscle precursor cells (MPCs) on iron-phosphate glass fibers

and b) MSCs differentiation on PBG fibers: Desmin (seen in green) is a cytoplasmic

marker of all skeletal muscle cells. Myogenin (seen as Red) is a nuclear marker of

differentiation. The Blue is DAPI(4,6-diamidino-2-phenylindole (which stains all

nuclei)). Reprinted from reference [165], Copyright (2004), with permission from

Elsevier.

3D PGF constructs (with composition 62.9P2O5–21.9Al2O3-15.2ZnO in mol % and

average diameter of 6.5 µm) were investigated for the construction of muscle organoid

units to resemble the structure of skeletal muscle cells [177]. PGF bundles were found

to be surrounded by fibrous sheaths and increasing cell numbers were observed on the

fiber construct due to the macro-topography of the 3D structure (see Figure 1.21).

Figure 1.21: a) Immunostaining shows a multinucleated myotube, expressing desmin

(green immunostaining) and sarcomeric actin (red immunostaining), attached to the

mesh scaffold on day 7, b) phase contrast micrograph of the myotube and mesh

scaffold, c) immunostaining demonstrating single myogenic desmin-positive cells

alongside gelatin coated glass ﬁbers, and d) phase contrast micrograph showing single

cells parallel to the glass ﬁbers. Reprinted from reference [177], Copyright (2005), with

permission from Elsevier.

PGFs with composition 50P2O5–30CaO–9Na2O–3SiO2–3MgO–(5−x)K2O–xTiO2 mol %

(where x = 0, 2.5, 5, respectively) were investigated for their influential role in neuronal

polarization and axonal growth direction by Vitale-Brovarone et al. [181]. It was

reported that aligned PGFs provided the directional cue for growing Dorsal Root

Ganglia (DRG) neuron cells along the fiber length (see Figure 1.22a). Additionally, the

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 active proliferation of neonatal olfactory bulb ensheathing cells (NOBEC) on PGFs was

seen to be extending along the fiber surface as shown in Figure 1.22b.

Figure 1.22: Confocal microscope images of a) DRG neurons on glass fibers presenting

long neuritis extended along the fiber axis direction and b) NOBEC cells showing active

proliferation on Phosphate glass fibers. Reprinted from reference [181], Copyright

(2012), with permission from Elsevier.

Moreover, degradation of PGFs in composite materials could lead to the formation of

micro-channels, which would play an important role in the perfusion and transportation

of nutrients, oxygen and formation of blood vessels. For example, Nazhat et al. [182]

investigated fomation of microchannels (30-40 µm diameter) within unidirectionally

orientated PGF reinforced collagen scaffolds (Figure 1.23). These spiral collagen-PGF

scaffolds were investigated as potential candidates for axonal outgrowth following

spinal cord injuries [183]. It was also observed that cylindrical scaffolds implanted into

transected spinal cords of rats exhibited better functional recovery compared to collagen

alone and no inﬂammatory responses were observed for both groups.

Figure 1.23: SEM images represent a) the cross-sectional PGF-collagen spiral

constructs and b) the formation of microchannels (indicated by arrows) via degradation

of the PGFs in within the ollagen spriral constructs. Reprinted with permission from

reference [182]. Copyright (2007) American Chemical Society.

Recently, PGF reinforced composites have also been investigated as fully resorbable

fracture fixation devices for load bearing applications [172]. Bioploymers (such as,

polylactic acid or polycaprolactone) reinforced with various compositions of PGF have

been fabricated in the form of rods (Figure 1.24a) and screws (Figure 1.24b) with

mechanical properties similar or superior to that of cortical bone. For example,

unidirectionally (UD) oriented PGF reinforced PLA composites exhibited an initial

flexural strength of 130 MPa and modulus of 11.5 GPa for a 24 % fiber volume fraction

composite [172]. In comparison, the mechanical properties of the human femur, tibia

and fibula were reported to have a modulus of elasticity and tensile strength in the range

of 10-22 GPa and 67-140 MPa, respectively [184,185].

Figure 1.24: Images of a) composite rod and b) composite screw prepared via the

forging process. Figure 1.22a) reprinted from reference [186], Copyright (2012), with

permission from Springer. Figure 1.22b) reprinted from reference [187], Copyright

(2013), with permission from Elsevier.

Polycaprolactone (PCL)-phosphate glass discs (see Figure 1.25a) were investigated in

vivo using a rat calvarium model [188]. No clinical complications were observed and

the lack of an inﬂammatory response showed that these composite materials were

biocompatible. Additionally, when compared to monolithic PCL discs the PCL-

phosphate glass discs revealed an increase in the amount of mineralized bone from 20 %

to 35 % over time. Extensive bone growth could also be seen after 26 weeks of

implantation which was characterized using histological examination (see Figure 1.25b)

[188].

Figure 1.25: (a) PCL-phosphate glass discs before implantation (8 mm diameter), and

(b) examples of bone associated with the dural face of implants after 26 weeks

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 implantation. Reprinted from reference [188], Copyright (2010), with permission from

John Wiley and Sons.

PCL-phosphate glass composites have also been explored as a root canal filling material

which was capable of sealing itself within the root canal within aqueous environments

+ 2+ 3− 4− 3− 5− whilst releasing certain ions (Na , Ca , PO4 , P2O7 , P3O9 , P5O10 ) at controlled

rates [189]. Compared to the conventional gutta-percha (GP) filler, these PCL-

phosphate glass composites exhibited better adaptation in the root canal and were found

to be firmly adhered to the canal wall.

Although a considerable amount of research has been undertaken on various

compositions of phosphate glasses in various geometries for biomedical applications,

there are still many challenges to overcome (for example, industrial-level scale-up fibre

manufacture) for commercial exploitation of these unique resorbable materials.

1.3.5 Electrical Properties

Phosphate glasses are not highly conducting materials, particularly oxide phosphate

glasses, like all other glass systems. However, there is a vast literature concerning basic

and applied research on the study of the electrical properties of phosphates, either of the

ionic or electronic conductivity type. Furthermore, there are some examples to which

phosphate-based glasses play a major role, such as LiPON [30], sulfide-phosphate

electrolytes [190], NASICON or NZP glass-ceramics [31,191] and cathode materials

[192]. Particularly, since the advent of all solid-state rechargeable batteries, the research

on novel glassy materials with high ionic or electronic conductivities has received

increasing interest.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 The application of glasses in either bulk or thin-film form may have several important

advantages for the practical configuration of the solid state battery. Glasses are

constituted by a single homogeneous phase without grain boundaries and the conduction

mechanism may take place in a much simpler way than in polycrystalline ceramics.

Furthermore, the composition of a glass can be tuned through small to large variations

without affecting the network structure and so the conductivity can be enhanced only by

adequate combinations of their chemistry. However, despite the wide range of

possibilities for the application of phosphate glasses in electrochemical systems, one

must always be concerned by the issue of chemical and electrochemical stability under

ambient conditions and in combination with the rest of components in the device, and in

this respect much remains to be done.

Lithium or sodium phosphate glasses might be thought to be the systems with the

highest ionic conductivity due to both alkali ions being very mobile. However,

experience has demonstrated that even though the content of alkali oxide may easily

reach 50 to 60 mol %, the room temperature conductivity remains at 10-6 S.cm-1, in the

best case, which is a value well below the conductivity of about 10-2 to 10-1 S.cm-1 of

the conventional liquid electrolytes. In a representative work by Martin et al. [193], the

conductivity of Li2O-P2O5 glasses was analyzed and compared with equimolar lithium

silicate and borate systems, and the authors concluded that the amount of charged non-

bridging oxygens in the network allows for, or facilitates, the fast migration of alkali

ions throughout a sort of channels of non-bridging oxygens, which in the case of

silicates is higher than in phosphates. Actually, this is also at the origin of the

interpretation for the increase of electrical conductivity in lithium phosphate glasses

after the nitrogen for oxygen substitution, as it will be seen below. The other cation that

gives rise to an abnormal very high conductivity is Ag+, and when combined with I-

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 anions in phosphate systems conductivity of up to 10-2 S.cm-1 can be attained, though

the stability of these glasses is even worse than that of the pure alkali phosphates and

their electrochemical window very narrow, thus finding a very limited use [194, 195].

It seems clear that either the mobility of the ionic species or their concentration cannot

be increased to the point necessary to reach conductivities at least of 10-4 S.cm-1, which

would be adequate for their application as solid electrolytes; therefore, some other

means should be addressed to further increase their room temperature conductivity. One

of these approaches concerns the polarizability of the anions that form the glass network

and that directly bond the modifier cations.

The two most well-known models that explain the ionic conduction in glasses are the

Anderson and Stuart model [196] and the theory of the weak electrolyte [197]. The

Anderson and Stuart model considers that the activation energy is composed of an

electrostatic term and a strain term and the total activation energy (F) for the conduction

of the mobile species may be written by equation (15):

2 2 F = zz0e / (r+r0) + 4GrD(r-rD) (15)

In the first term of the equation,  is a factor related to the finite displacement, z and z0

the valence of the ions,  a covalency parameter and r and r0 the interionic radii. This

first term is influenced by the covalent character of the bonds between the mobile

species, i.e. modifier ions, and the counterion to which cations are bonded. In this sense,

substitution of oxygen by a different anion, like S2-, establishing weaker bonds with the

modifier cations may reduce the associated electrostatic activation energy term and

therefore increase the conductivity of the glass. For instance, phosphate glasses

containing sulfide species, e.g. Li2S, are known to be highly conducting glasses and

have been of great interest in the last years [190]. In the second term, G refers to the

shear modulus and rD to a doorway that should allow the passing of an ion of radius r.

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 On the other hand, the weak electrolyte model assumes that not all modifier ions

participate of the conduction process and only a part of all cations will be mobile. In this

case, if one maintains the same anionic network, O2- in an oxide glass, and keeps

constant the total amount of modifier, it is possible to alter the number of mobile

species through compositional changes that increase the number of cations in the

energetic state favorable to take part of the electrical conduction process. This

approximation may be evidenced in mixed-former glasses, such boro-phosphate

compositions, in which the substitution of P2O5 by a second former oxide or

intermediate like B2O3 gives rise to a substantial increase of the electrical conductivity

of the glasses [53].

Electronic conduction in phosphate glasses has been studied in systems having mainly

Fe2O3 [198] and V2O5 [199]. Apart from their fundamental interest in understanding the

mechanisms of conduction in amorphous conductors, transition metal oxide containing

phosphate glasses have recently gained importance for their application as mixed ionic-

electronic conducting electrode materials in rechargeable batteries [192]. Nevertheless,

their room temperature conductivity remains quite low as compared to crystalline

electrode materials, generally below 10-10 S.cm-1, and most of the times glasses must be

processed in the form of glass-ceramics in order to gain further increase in their

conductivity, as shown by Garbarczyk et al. in V2O5-P2O5 glasses [200], or formulated

as mixed former glasses and with two or more transition metal oxides [201,202].

An interesting effect was postulated by Bazan et al. in alkali phosphate glasses having

tungstate oxide WO3, through which minima in the electrical conductivity are shown

when varying the lithium oxide content into a WO3-P2O5 glass system [203]. Such glass

compositions are able to exhibit both electronic and ionic conductivity, though

depending on the alkali content their behavior may be on the electronic dominant side or

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 in the ionic one. Furthermore, the reason for the observed minima upon alkali

incorporation was proposed to be due to a coupling between the ionic species and the

polarons, which for a certain amount of alkali oxide reduce the total conductivity down

to a minimum that constitutes a change of regime. The current study of mixed ionic-

electronic conducting phosphate glasses has been of importance for the application of

glasses in electrochromic devices [204].

Finally, it is worth mentioning that phosphate glasses have also received much attention

due to their particular ability for the retention of water and their relatively high proton

conductivities at moderate temperatures, an effect firstly observed by Abe et al. [205],

which makes them suitable as protonic conducting electrolytes in fuel cell systems

where polymeric systems fail under intermediate or high temperature operation

conditions [206].

Solid Electrolytes for Battery Applications

There is no doubt that nowadays the research on new electrochemical systems for the

storage and conversion of clean energy is of crucial interest for the development of

modern societies both at the small and large scale, and glasses may play a major role in

their progress. One of the most productive fields of research among the glass science

community has been the search for new solid electrolytes in the form of bulk glasses or

glass-ceramics and thin-films.

One of the most well-known commercialized solid state electrolytes with strong links to

phosphate glasses is the so-called LiPON, which is an amorphous thin-film material that

was developed as solid electrolyte for microbatteries and was firstly studied by J. B.

Bates and coworkers [30]. The layers of LiPON are grown by radio-frequency

magnetron sputtering techniques under N2 atmosphere, starting from Li3PO4 targets and

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 reaching an approximate composition of Li3.3PO3.8N0.22, depending on the processing

conditions. The authors soon realized that the electrochemical properties of the films as

well as their chemical and mechanical resistance were much improved with respect to

those of the lithium phosphates used before, as it happens after the nitrogen for oxygen

substitution in phosphate glasses. Furthermore, the layers of sputtered LiPON may serve

not only as single electrolyte layer but also as a means to protect batteries from the

deterioration occurring as due to the formation of dendrites at the metallic lithium

anodes when liquid electrolytes are employed [207]. Figure 1.26 shows a schematic

cross-section photograph of the typical configuration of a lithium microbattery where a

LiPON electrolyte is used, as shown by Bates et al. in reference [208].

Figure 1.26: Schematic cross-section of a thin-film lithium microbattery where LiPON

electrolyte can be seen in between the Li anode and cathode material. Reprinted from

reference [208], Copyright (2000), with permission from Elsevier.

In fact, the knowledge of LiPON electrolytes has been mostly based on the previous

existing research on nitrided phosphate glasses, and most authors have taken advantage

of the structural models in these glass systems for the elucidation of the structure and

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 properties changes in LiPON electrolytes. As seen in nitrided lithium metaphosphate

glasses, the study the sputtered LiPON layers has also shown to present the

dicoordinated (Nd) and tricoordinated nitrogen atoms (Nt) that were introduced above in

section 1.3.1. of Oxynitride Phosphate Glasses, and several authors have found that

there is an increase of the ionic conductivity of the films with the increase in the

proportion of Nt species of nitrogen [209]. Here, it should be pointed out that while the

conduction mechanisms might be strongly related between the one occurring in the film

LiPON electrolytes and that in the oxynitride lithium phosphate glass, the chemical

composition of both types of materials can be quite different and so the structural

interpretation that was before stated for the bulk glass systems should not be strictly

used in the structure-properties relationship in the LiPON films [210]. The ionic

conductivity change by N/O substitution in lithium phosphate glasses has been studied

thoroughly in the last years by F. Muñoz and coworkers. In a first work, the authors

showed that independent of the modifier content in nitrided Li2O-P2O5 glasses, the room

temperature conductivity increases with nitrogen up to values of N/P ratio about 0.2-0.3,

then showing a much lower or null variation with further nitrogen addition [211]. In that

work, they proposed that the increase in electrical conductivity should be due to the fact

that nitridation produces a decrease of the bridging to non-bridging oxygens ratio

(BO/NBO) giving rise to a larger net amount of NBOs as the nitrogen content increases,

Thus, the nitrogen for oxygen substitution creates conduction paths favorable for the

ionic hopping mechanism of lithium and this hypothesis was later evidenced by

Mascaraque et al. through the study of the BO/NBO ratio of a range of oxynitride glass

compositions with varying Li2O/P2O5 ratios [212].

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Phosphates have also been studied in the form of glass-ceramics trying to reproduce the

composition of crystalline compounds based on lithium titanium phosphates with the

NASICON (sodium superionic conductor) type structure, or NZP, in NaZr2P3O12. The

systems studied have a variety of alternatives but most are formulated within Li2O-

Al2O3-MO2-P2O5, where M can be Ge or Ti and all belong to the ortho-phosphate

composition [31]. They are prepared through a melt-quenching method at temperatures

from 1400 to 1500ºC, and then heat treated in order to develop the LATP (Lithium

Aluminum Titanium Phosphate) crystalline phases, giving rise normally to a fully

crystallized material. In fact, the melts have a very high tendency for spontaneous

crystallization given the very low amount of P2O5 from which they are formulated,

which in many cases originates in inhomogeneous glass/glass-ceramics. However, they

can easily reach room temperature conductivities in the order of 10-4 S.cm-1, which

allow them to be used as solid-state electrolytes in their bulk form. Recently, LATP

glass-ceramics have been tested into lithium-air secondary batteries as a protecting layer

of the lithium metallic electrode or even in combination with LiPON electrolyte

[213,214].

1.3.6 Phosphate Glasses for Waste Storage

As it has been discussed in previous sections, phosphate glasses usually have a chemical

durability which is inferior to that of most silicate and borosilicate glasses. The

dissolution rate is very sensitive to phosphate glass compositions, and widely ranges

from over 10-4 to almost 10-9 g.cm-2.min-1. In order to improve the chemical durability,

cations with high electrostatic field, like Zn2+ and Pb2+ are used to increase greatly the

covalence of the P–O–M bonds. Trivalent cations (Fe3+, Al3+) have also been

successfully introduced to strengthen the glass network reticulation. However, the iron

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 phosphate glasses are an exception. Binary and ternary iron phosphate glasses

containing more than 25 mol % Fe2O3 have an exceedingly good chemical durability.

Their dissolution rate at 90°C in distilled water or in saline solution is up to 100 times

lower than that of window glass [215]. In addition to their excellent chemical durability,

iron phosphate glasses can usually be melted between 950°C and 1150°C in only a few

hours [216] since the melts are fluid and rapidly homogenize.

Vitrification of high level nuclear wastes (HLW) has received a great worldwide

attention since more than 40 years. The associated technology, which has reached its

maturity, presents several relative advantages such as the ease of production, the

insensitivity to the waste composition fluctuations and the high resistance to potential

alterations due to the combined effects of heat, radiation and aqueous solutions.

Many vitrification processes have been developed and tested. They can be classified in

two types: discontinuous and continuous, with a clear tendency in favor of the last one.

Each of them includes three temperature dependent phases: drying, calcining and

vitrification. HLW can be incorporated directly into the glass network or by glass

encapsulation in the form of a composite material. Depending on the process chosen for

the large scale vitrification plant, discontinuous pouring of the melt can be achieved to

fill a canister, or the molten glass can be separated into droplets allowing the continuous

production of beads which are poured in a canister and then embedded in a molten lead

matrix (Vitromets). In each case, after filling, the storage containers are welded, cooled

down and then transported to a storage facility.

Among several hundred of glass frits described in literature, only a limited number of

glass compositions is used on a large scale or considered as reference materials. Table

VI illustrates, for most countries, the glass compositions used for HLW incorporation

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 and the corresponding waste loading limit. These glass formulations may be subject to

slight changes depending on the fluctuations in the waste composition.

Table VI: Compositions of some nuclear waste glasses in wt. %.

Waste

Glass/country SiO2 P2O5 B2O3 Al2O3 CaO MgO Na2O Misc. loading

(%)

France 47.2 - 14.9 4.4 4.1 - 10.6 18.8 ≤ 28

USA 49.8 - 8.0 4.0 1.0 1.4 8.7 27.1 ≤ 33

UK 47.2 - 16.9 4.8 - 5.3 8.4 17.4 ≤ 25

Germany/ 52.7 - 13.2 2.7 4.6 2.2 5.9 18.7 ≤ 30 Belgium

Russia - 52 - 19.0 - - 21.2 7.8 ≤ 10

High waste loading are achieved both in borosilicate and alumino-phosphate glasses,

however molten phosphate glasses are known to be highly corrosive to refractory liners,

contrary to the borosilicate melts. This behavior is probably the main reason why their

application has not been envisaged for waste from nuclear fuel reprocessing in most

countries excepted in Russia since 1987. Those chemical compositions are the results of

a compromise between glass durability and technical feasibility. Attention has to be

paid to the processing temperatures and the viscosity of the melt which should be the

lowest as possible, 1150°C and 10 Pa.s, respectively. Thus, the volatility of fission

products, like Cs, could be avoided and a control of the pouring into the canisters could

be ensured, minimizing the blending problems. The critical characteristics of the glass

waste are the possibility to undergo phase transformations during cooling, thus

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 degrading the mechanical integrity, and the radiation effects on any of the properties of

the glass and especially on its chemical durability. The reintroduction of radionuclides

in the biosphere is considered as fulfilled by water leaching. It is the only conceivable

way to transport radioactive materials away from a waste repository. The release rates

of radionuclides can be expressed on the basis of normalized dissolution rate (NDR)

according to various test protocols such as ISO 6961-1982 and supposing that the

evaluation of the long-term performance of glass wastes (i.e. more than 10000 years)

can be inferred to laboratory time scale experiments. Typical NDR values of glass

wastes are of 10-6 g.cm-2.day-1. However, such a value can be dramatically increased, by

several orders of magnitude, when the glass has undergone crystallization of water

soluble phases during cooling.

The following table summarizes typical data of HLW borosilicate and phosphate glasses

[217].

Table VII: Typical properties of HLW glasses

Compressive NDR (28th day) Glass Density (g.cm-3) Thermal stability Strength (MPa) (10-6. g.cm-2.day-1)

Borosilicate 2.7 22 - 54 0.3 (Cs) - 0.2 (Sr) ≥ 550°C

Phosphate 2.6 9 - 14 1.1 (Cs) - 0.4 (Sr) ≥ 450°C

Depending on their origin, HLW can contain phosphates (up to 15 wt. % P2O5), iron

oxide (up to 25 wt. %) and other heavy metal oxides such as Bi2O3 (up to 30 wt. %) or

UO2 (up to 30 wt. %). As a consequence, the borosilicate glass used for the vitrification

of the nuclear wastes can present a phase separation below Tg resulting from a liquid

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 immiscibility above Tg. This is due to a P2O5 content larger than a critical concentration,

which is clearly dependent on the glass composition and vary from 0.5 to 7 wt. % [216].

This phase separation may also induce a mismatch in the coefficients of thermal

expansion that results in local stresses and cracks in the glassy phase increasing the risk

of failure of the glass waste form.

To avoid this problem, two possibilities may be considered. The easiest is to dilute such

HLW’s in order to decrease the P2O5 content below the phase separation threshold. This

unfortunately leads to an increase of the volume of the resulting waste form.

An alternative is to take advantage of the presence of phosphorous and iron oxides in

these HLW and to produce a glassy iron phosphate waste form. It has been

demonstrated that the melting and fining process of such iron phosphate glasses can be

achieved in less than 2 h at temperatures which are usually lower than those needed for

borosilicate glasses. Furthermore, these glasses exhibit chemical durability (DR) values

of about 10-9 g.cm-2.day-1, evaluated from the weight loss of bulk samples immersed in

deionized water at 90°C. These values are similar to that of a CVS-IS standard

borosilicate glass made by Pacific Northwest National Labs [216]. Furthermore, the

lowest DR values are obtained for glasses with an O/P atomic ratio close to 3.5. From a

structural point of view, it corresponds to pyrophosphate P2O7 groups which are bonded

together by iron ions. Assuming that the corrosion of phosphate glasses occurs by the

hydration of P-O-P bonds, the excellent chemical durability of these glasses may be

attributed to the replacement of the P-O-P bonds by the more hydration resistant Fe-O-P

bonds.

Using the same approach, a simulated sodium bearing waste (SBW) was successfully

vitrified in iron phosphate glasses at a maximum waste loading of 40 wt. % using both

conventional and cold crucible induction melting techniques [218]. No sulfate

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 segregation or crystalline phases were detectable. The wasteforms containing 40 wt. %

SBW satisfy current requirements for aqueous chemical durability. The fluid

wasteforms can be melted at a relatively low temperature (1000°C) and for short times

(<6 h). These properties combined with a significantly higher waste loading and the

feasibility of cold crucible induction melting, offer considerable savings in time, energy,

and cost for vitrifying the SBW.

Finally, before 1990, the nuclear waste management community considered borosilicate

glasses as the “one size fits all” host matrix. As the diversity and chemical complexity

of the nuclear wastes became more apparent and as the range of nuclear wastes tested in

borosilicate glasses widened, the potential problems of the incompatibility of certain

wastes with borosilicate glasses became more evident. A logical solution to this

problem is to have alternative host matrices that are better suited for vitrifying such

problematic wastes and match the glass to the waste.

Conclusions and perspectives

Phosphate based glasses have been investigated as one of the most versatile systems

which have found application in many different technological fields, from ionic

conductors to laser host materials, sealing glasses, biomaterials and even as matrices for

immobilization of toxic wastes. Their particular thermal characteristics make them

unique for certain applications where other glass types would not be appropriate, and

despite their relatively low chemical durability, normally associated due to their

hygroscopic character, several methods have successfully been applied to address this

drawback. Furthermore, their ability to incorporate transition metal elements, rare-earths

or heavy metals, as well as for the substitution of oxygen by a number of different

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 elements (N, S, F, I,…), significantly increases the applicability of phosphate glass

systems, giving rise to new, improved and enhanced properties constantly.

Furthermore, structural studies of phosphate glasses have provided very precise

representations of their atomic arrangements, not only at the local scale but also at

longer range orders, enabling further insight into their structure-property relationships.

Therefore, phosphate glasses have a very bright future ahead with further new

developments transpiring into further new applications. As these new developments

come to the fore, their formulations will inevitably become more complex and

elucidation of their properties based on their structure will remain of key importance,

for which novel modelling techniques of amorphous systems may well be required and

become a central issue.

Acknowledgements

F. Muñoz thanks funding from projects MAT2013-48246-C2-1-P from MINECO of

Spain and I-link+0959 from CSIC. I. Ahmed would like to acknowledge the Faculty of

Engineering, Advanced Materials Research Group, University of Nottingham, for

provision of studentship funds.

[1] W. Vogel: Glass Chemistry (Springer-Verlag, Berlin, Heidelberg 1994).

[2] N.N. Greenwood, A. Earnshaw: Chemistry of the Elements (Pergamon Press, Oxford

1984).

[3] V.M. Goldschmidt: Geochemische Verteilungsgesetze der Elemente, Skr. Nor.

Vidensk Akad. K1, 1; Mat. Naturvidensk. K1 8, 7-156 (1926).

[4] W.J. Zachariasen: The atomic arrangement in Glass, J. Am. Ceram. Soc. 54, 3841-

3851 (1932).

[5] A. Dietzel: Die kationenfeldstärken und irhe beziehungen zu entglasungsvorgängen,

zur berbindungsbildung und zu den schmelzpunkten von silicaten, Z. Elektrochem. 48,

9-23 (1942).

[6] E. Thilo: Die Kondensierten Phosphate, Die Naturwissenschaften 46(11), 367-373

(1959).

[7] J.R. Van Wazer: Phosphorus and its compounds vol. I (Interscience, New York

1958).

[8] R.K. Brow, D.R. Tallant, J.J. Hudgens, S.W. Martin, A.D. Irwin: The short-range

structure of sodium ultraphosphate glasses, J. Non-Cryst. Solids 177, 221-228 (1994).

[9] U.W. Hoppe: A structural model for phosphate glasses, J. Non-Cryst. Solids 195,

138-147 (1996).

[10] J.R. Jones, A.G. Clare: Bio-Glasses: An Introduction (John Wiley & Sons, United

Kingdom 2012).

[11] M. Ren, S. Cai, W. Zhang, T. Liu, X. Wu, P. Xu, D. Wang: Preparation and

chemical stability of CaO-P2O5-Na2O-B2O3 porous glass-ceramics, J. Non-Cryst. Solids

380, 78-85 (2013).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [12] J.H. Campbell, J.S. Hayden, A. Marker: High Power Solid-State Lasers: a Laser

Glass Perspective, Int. J. App. Glass Sci. 2 [1], 3-29 (2011).

[13] I. Ahmed, A.J. Parsons, G. Palmer, J.C. Knowles, G.S. Walker, C.D. Rudd: Weight

loss, ion release and initial mechanical properties of a binary calcium phosphate glass

fibre/PCL composite, Acta Biomaterialia 4, 1307-1314 (2008).

[14] J.C. Knowles: Phosphate based glasses for biomedical applications, Journal of

Materials Chemistry 13, 2395-2401 (2003).

[15] I. Ahmed, M. Lewis, I. Olsen, J.C. Knowles: Phosphate glasses for tissue

engineering: Part 2. Processing and characterisation of a ternary-based P2O5-CaO-Na2O

glass fibre system, Biomaterials 25, 501-7 (2004).

[16] I. Ahmed, M. Lewis, I. Olsen, J.C. Knowles: Phosphate glasses for tissue

engineering: Part 1. Processing and characterisation of a ternary-based P2O5–CaO–Na2O

glass system, Biomaterials 25, 491-499 (2004).

[17] K.M.Z. Hossain, M.S. Hassan, R. Felfel, If. Ahmed: Development of phosphate-

based glass fibers for biomedical applications, in Hot Topics in Biomaterials, 104-115

(Future Science Ltd., London 2014).

[18] M.J. Weber: Science and technology of laser glass, J. Non-Cryst. Solids 123, 208-

222 (1990).

[19] J.H. Campbell, T.I. Suratwala: Nd-doped phosphate glasses for high-energy/high-

peak-power lasers, J. Non-Cryst. Solids 263&264, 318-341 (2000).

[20] T.T. Fernandez, P. Haro-González, B. Sotillo, M. Hernandez, D. Jaque, P.

Fernandez, C. Domingo, J. Siegel, J. Solis: Ion migration assisted inscription of high

refractive index contrast waveguides by femtosecond laser pulses in phosphate glass,

Optics Letters 38, 5248-5251 (2013).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [21] J. del Hoyo, R. Martinez Vazquez, B. Sotillo, T.T. Fernandez, J. Siegel, P.

Fernández, R. Osellame, J. Solis: Control of waveguide properties by tuning

femtosecond laser induced compositional changes, App. Phys. Lett. 105, 131101

(2014).

[22] N. Marquestaut, Y. Petit, A. Royon, P. Mounaix, T. Cardinal, L. Canioni: Three-

Dimensional Silver Nanoparticle Formation Using Femtosecond Laser Irradiation in

Phosphate Glasses: Analogy with Photography, Adv. Funct. Mater. 24, 5824–5832

(2014)

[23] J.E. Shelby: Introduction of Glass Science and Technology 2nd Ed. (The Royal

Society of Chemistry, Cambridge 2005).

[24] R. Morena: Phosphate glasses as alternatives to Pb-based sealing frits, J. Non-

Cryst. Solids 263&264, 382-387 (2000).

[25] M. Rajaram, D.E. Day: Preparation and properties of oxynitride phosphate glasses

made from 27R2O20BaO3Al2O350P2O5, J. Non-Cryst. Solids 102, 173-180 (1988).

[26] M. Duclot, J.-L. Souquet: Glassy materials for lithium batteries: electrochemical

properties and devices performances, J. Power Sources 97-98, 610-615 (2001).

[27] T. Ishiyama, S. Suzuki, J. Nishii, T. Yamashita, H. Kawazoe, T. Omata: Proton

conducting tungsten phosphate glass and its application in intermediate temperature fuel

cells, Solid State Ionics 262, 856-859 (2014).

[28] T. Ishiyama, J. Nishii, T. Yamashita, H. Kawazoe and T. Omata: Electrochemical

substitution of sodium ions with protons in phosphate glass to fabricate pure proton

conducting glass at intermediate temperatures, J. Mater. Chem. A 2, 3940-3947 (2014).

[29] Satoshi Nakata, Takuya Togashi, Tsuyoshi Honma, Takayuki Komatsu: Cathode

properties of sodium iron phosphate glass for sodium ion batteries, Journal of Non-

Crystalline Solids 450, 109-115 (2016).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [30] X. Yu, J.B. Bates, G.E. Jellison Jr., F.X. Hart: A Stable Thin‐Film Lithium

Electrolyte: Lithium Phosphorus Oxynitride, J. Electrochem. Soc. 144 [2], 524-532

(1997).

[31] X. Xu, Z. Wen, Z. Gu, X. Xu, Z. Lin: Lithium ion conductive glass-ceramics in the

system Li1.4Al0.4(Ge1-xTix)1.6(PO4)3 (x=0-1.0), Solid State Ionics 171, 207-213 (2004).

[32] S.T. Reis, M. Karabulut, D.E. Day: Structural features and properties of lead-iron-

phosphate nuclear wasteforms, J. Nucl. Mat. 304, 87-95 (2002).

[33] J.R. Van Wazer: Structure and Properties of the Condensed Phosphates. II. A

Theory of the Molecular Structure of Sodium Phosphate Glasses, J. Am. Chem. Soc. 72,

644-647 (1950).

[34] E. Lippmaa, M. Maegi, A. Samoson, G. Engelhardt, A. Grimmer: Structural studies

of silicates by solid-state high-resolution silicon-29 NMR, J. Am. Chem. Soc. 102 (15),

4889-4893 (1980).

[35] R.K. Brow, Review: the structure of simple phosphate glasses, J. Non-Cryst. Solids

263&264, 1-28 (2000).

[36] B.H. Jung, D.N. Kim, H.-S. Kim: Properties and structure of (50-x)BaO-xZnO-

50P2O5 glasses, J. Non-Cryst. Solids 351, 3356-3360 (2005).

[37] L. Muñoz-Senovilla, F. Muñoz: Behaviour of viscosity in metaphosphate glasses,

J. Non-Cryst. Solids 385, 9-16 (2014).

[38] S. Mamedov, D. Stachel, M. Soltwitsch, D. Quitmann: Local environment and

dynamics of PO4 tetrahedra in Na-Al-PO3 glasses and melts, J. Chem. Phys. 123,

124515-124527 (2005).

[39] R.J. Kirkpatrick, R.K. Brow: Nuclear magnetic resonance investigation of the

structures of phosphate and phosphate-containing glasses: a review, Solid State Nucl.

Magn. Res. 5, 9-21 (1995).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [40] R.K. Brow, D.R. Tallant, S.T. Myers, C.C. Phifer: The short-range structure of zinc

polyphosphate glass, J. Non-Cryst. Solids 191, 45-55 (1995).

[41] F. Muñoz, F. Agulló-Rueda, L. Montagne, R. Marchand, A. Durán, L. Pascual:

Structure and properties of (25-x/2)Li2O-(25-x/2)Na2O-xPbO-50P2O5 metaphosphate

glasses, J. Non-Cryst. Solids 347, 153-158 (2004).

[42] T.M. Duncan, D.C. Douglass: On the 31P chemical shift anisotropy of condensed

phosphates, Chemical Physics 87, 339-349 (1984).

[43] M. Feike, R. Graf, I. Schnell, C. Jäger, C.W. Spiess: Structure of crystalline

phosphates from 31P Double-Quantum NMR spectroscopy, J. Am. Chem. Soc. 118,

9631-9634 (1996).

[44] P. Rajbhandari, Y. Chen, B. Doumert, L. Montagne, G. Tricot: Investigation of

zinc alkali pyrophosphate glasses. Part II : Local and medium range orders analysed by

1D/2D NMR, Mat. Chem. Phys. 155, 23-29 (2015).

[45] R. Witter, P. Hartmann, J. Vögel, C. Jäger: Measurements of chain length

distributions in calcium phosphate glasses using 2D 31P double quantum NMR, Solid

State Nucl. Magn. Res. 13, 189-200 (1998).

[46] F. Fayon, G. Le Saout, L. Emsley, D. Massiot : Through-bond phosphorus-

phosphorus connectivities in crystalline and disordered phosphates by solid-state NMR,

Chem. Comm. 1702-1703 (2002).

[47] F. Fayon, I.J. King, R.K. Harris, J.S.O. Evans, D. Massiot: Application of the

through-bond correlation NMR experiment to the characterization of crystalline and

disordered phosphates, C. R. Chimie 7, 351-361 (2004).

[48] J. Ren, H. Eckert: Applications of DQ-DRENAR for the structural analysis of

phosphate glasses, Solid State Nucl. Magn. Res. 72, 140-147 (2015).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [49] P. Guerry, M. E. Smith, S.P. Brown: 31P MAS Refocused INADEQUATE Spin-

Echo (REINE) NMR Spectroscopy: Revealing J Coupling and Chemical Shift Two-

Dimensional Correlations in Disordered Solids, J. Am. Chem. Soc. 131, 11861-11874

(2009).

[50] M. Lahaye, B. Doumert, B. Revel, K. Ben Tayeb, H. Vezin, G. Tricot : Application

of Magnetic Resonance Spectroscopies to the xZnO-(100-x)NaPO3 Glass System :

Glass Network Organization and Effect of Co2+ doping, J. Phys. Chem. C 119, 17288-

17297 (2015).

[51] F. Moreau, A. Durán, F. Muñoz : Structure and properties of high Li2O-containing

aluminophosphate glasses, J. Eur. Ceram. Soc. 29, 1895-1902 (2009).

[52] L. Van Wüllen, G. Tricot, S. Wegner: An advanced NMR protocol for the

structural characterization of aluminophosphate glasses, Solid State Nucl. Mag. Res. 32,

44-52 (2007).

[53] F. Muñoz, L. Montagne, L. Pascual, A. Durán : Composition and structure effects

on the properties of lithium borophosphate glasses showing boron anomaly, J. Non-

Cryst. Solids, 355, 2571-2577 (2009).

[54] M. Zeyer, L. Montagne, V. Kostoj, G. Palavit, D. Prochnow, C. Jaeger : 17O

nuclear magnetic resonance study of Na2O-P2O5 glasses, 311, 223-232 (2002).

[55] A. Flambard, L. Montagne, L. Delevoye : A new 17O-isotopic enrichment method

for the NMR characterisation of phosphate compounds, Chemm. Commun. 3426-3428

(2006).

[56] A. Flambard, L. Montagne, L. Delevoye, G. Palavit, J.-P. Amoreaux, J.-J. Videau :

Solid-state NMR study of mixed-network sodium-niobium phosphate glasses,

345&346, 75-79 (2004).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [57] R.E. Dinnebier, S.J.L. Billingue (Eds.): Powder diffraction: theory and practice

(RSC publishing, Cambridge 2008).

[58] U. Hoppe, G. Walter, R. Kranold, D. Stachel: Structural specifics of phosphate

glasses probed by diffraction methods: a review, J. Non-Cryst. Solids 263&264, 29-47

(2000).

[59] J. Biscoe, A.G. Pincus, C.S. Smith, B.E. Warren: X-ray study of lime-phosphate

and lime-borate glass, J. Am. Ceram. Soc. 24, 116-119 (1941).

[60] Y. Waseda: The Structure of Non-Crystalline Materials (McGraw-Hill, New York,

1980).

[61] U. Hoppe, D. Stachel, D. Beyer : Oxygen coordination of metal ions in phosphate

and silicate glasses studied by a combination of x-ray and neutron diffraction, Phys.

Scripta T57, 122-126 (1995).

[62] E. Matsubara, K. Sugiyama, Y. Waseda, M. Ashizuka, E. Ishida: Structural

analysis of zinc metaphosphate glass by anomalous X-ray scattering, J. Mater. Sci. Lett.

9, 14-16 (1990).

[63] A. Balerna, M. Bionducci, A. Falqui, G. Licheri, C. Meneghini, G. Navarra, M.

Bettinelli : A structural study of Sr metaphosphate glass by anomalous X-ray scattering

and EXAFS spectroscopy, J. Non-Cryst. Solids 232&234, 607-612 (1998).

[64] P.H. Gaskell, J. Zhao, P. Boden, P. Chieux: Structure of a copper sodium

phosphate glass by neutron scattering with isotopic substitution, J. Non-Cryst. Solids

150, 80-86 (1992).

[65] U. Hoppe, E. Metwalli, R. K. Brow, J. Neuefeind: High-energy X-ray diffraction

study of La co-ordination in lanthanum phosphate glasses, J. Non-Cryst. Solids 297,

263-274 (2002).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [66] U. Hoppe, D. Ilieva, J. Neuefeind: The structure of gallium phosphate glasses by

high-energy X-ray diffraction, Zeitschrift fur Naturforschung A 57, 709-715 (2002).

[67] U. Hoppe, M. Karabulut, E. Metwalli, R. K. Brow, P. Jovari: The Fe-O

coordination in iron phosphate glasses by X-ray diffraction with high energy photons, J.

Phys. Condensed Matt. 15, 6143–6153 (2003).

[68] U. Hoppe, R. Kranold, A. Ghosh, C. Landron, J. Neuefeind, P. Jovari;

Environments of lead cations in oxide glasses probed by X-ray diffraction, J. Non-

Cryst. Solids 328, 146-156 (2003).

[69] U. Hoppe, Y. Dimitriev, P. Jovari: Structure of zinc phosphate glasses of 75 and 80

mol % ZnO content studied by X-ray diffraction and reverse Monte Carlo simulations,

Zeitschrift fur Naturforschung A 60, 517-526 (2005).

[70] U. Hoppe, R. K. Brow, B. C. Tischendorf, A. Kriltz, P. Jóvári, A. Schöps, A.C.

Hannon: Structure of titanophosphate glasses studied by X-ray and neutron diffraction,

J. Non-Cryst. Solids 353, 1802-1807 (2007).

[71] U. Hoppe, G. Walter, A. Barz, D. Stachel, A.C. Hannon: The P-O bond lengths in

vitreous P2O5 probed by neutron diffraction with high real-space resolution, J. Phys.:

Condens. Matt. 10, 261-270 (1998).

[72] K. Suzuki, M. Ueno: Experimental discrimination between bridging and non-

bridging oxygen phosphorus bonds in P2O5-Na2O glass by pulsed neutron total

scattering, Journal de Physique 46 (C8), 261-265 (1985).

[73] U. Hoppe, G. Walter, R. Kranold, D. Stachel: An X-ray diffraction study of the

structure of vitreous P2O5, Z. Naturforsch. A53, 93-04 (1998).

[74] R. Gresch, W. Müller-Warmuth, H. Dutz: X-ray photoelectron spectroscopy of

sodium phosphate glasses, J. Non-Cryst. Solids 34, 127-136 (1979).

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [75] P. Losso, B. Schnabel, C. Jäger, U. Sternberg, D. Stachel, D.O. Smith: 31P NMR

investigations of binary alkaline earth phosphate glasses of ultraphosphate composition,

J. Non-Cryst. Solids 143, 265-273 (1992).

[76] R.K. Brow, C.C. Phifer, G.L. Turner, R.J. Kirkpatrick: Cation effects on 31P MAS

NMR chemical shifts of metaphosphate glasses, J. Am. Ceram. Soc. 74, 1287-1290

(1991).

[77] G.N. Greaves, S.J. Gurman, L.F. Gladden, C.A. Spence, B.C. Sales, L.A. Boatner,

R.N. Jenkins: A structural basis for the corrosion resistance of lead-iron-phosphate

glasses: An X-ray absorption spectroscopy study, Philos. Mag. B 58, 271-283 (1988).

[78] U. Hoppe, G. Walter, D. Stachel, A.C. Hannon: Short-range order details of

metaphosphate glasses studied by pulsed neutron scattering, Z. Naturforsch. A50, 684-

692 (1995).

[79] U. Hoppe, G. Walter, D. Stachel, A.C. Hannon. Short-range order in KPO3 glass

studied by neutron and X-ray diffraction, Z. Naturforsch. A51, 179-186 (1996).

[80] T. Uchino, Y. Ogata: Ab-initio molecular orbital calculations on the electronic

structure of phosphate glasses. Binary alkali metaphosphate glasses, J. Non-Cryst.

Solids 191, 56-70 (1995).

[81] U. Hoppe, G. Walter, D. Stachel: The short range order of metaphosphate glasses

investigated by X-ray diffraction, Phys. Chem. Glasses, 33, 216-221 (1992).

[82] U. Hoppe, G. Walter, R. Kranold, D. Stachel, A. Barz: The dependence of

structural peculiarities in binary phosphate glasses on their network modifier content, J.

Non-Cryst. Solids 192&193, 28-31 (1995).

[83] J.J. Hudgens, S.W. Martin: Glass transition and infrared spectra of low alkali,

anhydrous lithium phosphate glasses, J. Am. Ceram. Soc. 76, 1691-1696 (1993).

100

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [84] J.J. Hudgens: The structure and properties of anhydrous alkali ultra-phosphate

glasses, (PhD Thesis, Iowa State University, Ames 1994).

[85] U. Hoppe, R. Kranold, D. Stachel, A. Barz, A.C. Hannon: A neutron and X-ray

diffraction study of the structure of the LaP3O9 glass, J. Non-Cryst. Solids 232&234,

44-50 (1998).

[86] U. Hoppe, R. Kranold, D. Stachel, J. Neuefeind: Oxygen coordination of modifier

cations in metaphosphate glasses probed by high energy X-ray diffraction, Phosphorus

Res. Bull. 10, 546-551 (1999).

[87] A. Musinu, G. Paschina, G. Piccaluga, G. Pinna: Short range order of

metaphosphate glasses by X-ray diffraction, J. Non-Cryst. Solids 177, 97-102 (1994).

[88] S.C. Moss, D.L. Price, in: D. Adler, H. Fritzsche, S.R. Ovshinsky (Eds.), Physics of

Disordered Materials, (Plenum, New York 1985).

[89] S.R. Elliott : Extended-range order, interstitial voids and the first sharp diffraction

peak of network glasses, J. Non-Cryst. Solids 182, 40-48 (1995).

[90] P.H. Gaskell, D.J. Wallis: Medium-range order in silica, the canonical network

glass, Phys. Rev. Lett. 76, 66-69 (1996).

[91] G. Walter, U. Hoppe, T. Baade, R. Kranold, D. Stachel: Intermediate-range order

in MeO-P2O5 glasses, J. Non-Cryst. Solids 217, 299-307 (1997).

[92] K. Suzuya, D.L. Price, C.-K. Loong, S.W. Martin: Structure of vitreous P2O5 and

alkali phosphate glasses, J. Non-Cryst. Solids 232&234, 650-657 (1998).

[93] K. Suzuya, D.L. Price, C.-K. Loong, S. Kohara: The structure of magnesium

phosphate glasses, J. Phys. Chem. Solids 60, 1457-1460 (1999).

[94] K. Suzuya, K. Itoh, A. Kajinami, C.-K. Loong: The structure of binary zinc

phosphate glasses, J. Non-Cryst. Solids 345&346, 80-87 (2004).

101

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [95] H. Rietveld: A profile refinement method for nuclear and magnetic structures, J.

Appl. Crystallogr. 2, 65-71 (1969).

[96] R. L. McGreevy and L. Pusztai: Reverse Monte Carlo simulation: A new technique

for the determination of disordered structures, Mol. Simul. 1, 359-367 (1988).

[97] R. L. McGreevy: Reverse Monte Carlo modelling, J. Phys.: Condens. Matter 13,

R877-R914 (2001).

[98] B. J. Alder and T. E. Wainwright: Studies in molecular dynamics. I. General

method, J. Chem. Phys. 31, 459-466 (1959).

[99] A. N. Cormack and Y. Cao: Molecular dynamics simulation of silicate glasses,

Molecular Engineering 6, 183-227 (1996).

[100] A. K. Soper: Test of the empirical potential structure refinement method and a

new method of application to neutron diffraction data on water, Mol. Phys. 99, 1503-

1516 (2001).

[101] A. K. Soper: Partial structure factors from disordered materials diffraction data:

an approach using empirical potential structure refinement, Phys. Rev. B 72, 104204-

104216 (2005).

[102] K. M. Wetherall, D. M. Pickup, R. J. Newport and G. Mountjoy: The structure of

calcium metaphosphate glass obtained from x-ray and neutron diffraction and reverse

Monte Carlo modelling, J. Phys.: Condens. Matter 21, 035109 (2009).

[103] A. Tilocca, Models of structure, dynamics and reactivity of bioglasses: a review,

J. Mater. Chem. 20, 6848-6858 (2010).

[104] Y. Shaharyar, E. Wein, J.-J. Kim, R. E. Youngman, F. Muñoz, H.-W. Kim, A.

Tiloccah, A. Goel, Structure-solubility relationships in fluoride-containing phosphate

based bioactive glasses, J. Mater. Chem. B 3, 9360-9373 (2015).

102

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [105] J.K. Christie, R.I. Ainsworth, N.H. de Leeuw, Ab initio molecular dynamics

simulations of structural changes associated with the incorporation of ﬂuorine in

bioactive phosphate glasses, Biomaterials 35, 6164-6171 (2014).

[106] E.A. Ruben, M.S. Chapman, J.D. Evanseck: Hydrogen Bonding Mediated by

Chemical Interactions Determines Hydration Enthalpy Differences of Phosphate Water

Clusters, J. Phys. Chem. A 111, 10804-10814 (2007).

[107] F. Delahaye, L. Montagne, G. Palavit, J.C. Touray, P. Baillif: Acid dissolution of

sodium-calcium metaphosphate glasses, J. Non-Cryst. Solids 242, 25-32 (1998).

[108] D.E. Day, Z. Wu, C.S. Ray, P. Hrma: Chemically durable iron phosphate glass

wasteforms, J. Non-Cryst. Solids 241, 1-12 (1998).

[109] B.C. Bunker, G.W. Arnold, J.A. Wilder: Phosphate glass dissolution in aqueous

solutions, J. Non-Cryst. Solids 64, 291-316 (1984).

[110] H. Gao, T. Tan, D. Wang: Dissolution mechanism and release kinetics of

phosphate controlled release glasses in aqueous medium, J. Controlled Release 96, 29-

36 (2004).

[111] F. Döhler, A. Mandlule, L. Van Wüllen, M. Friedrich, D.S. Brauer: 31P NMR

characterization of phosphate fragments during dissolution of calcium sodium

phosphate glasses, J. Mat. Chem. B 3, 1125-1134 (2015).

[112] P.E. Gray, L.C. Klein: The chemical durability of sodium ultraphosphate glasses,

Glass Technology 24 [4], 202-206 (1983).

[113] N. Mascaraque, A. Durán, F. Muñoz: Effect of fluorine and nitrogen on the

chemical durability of lithium phosphate glasses, J. Non-Cryst. Solids 417-418, 60-65

(2015).

[114] H. Takebe, Y. Baba, M. Kuwabara: Dissolution behavior of ZnO-P2O5 glasses in

water, J. Non-Cryst. Solids 352, 3088-3094 (2006).

103

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [115] R. Marchand: Mise en évidence de verres de phosphates contenant de l’azote,

C.R. Acad. Sc. Paris 294, 91-94 (1982).

[116] S. Hampshire, M.J. Pomeroy: Grain boundary glasses in silicon nitride : A review

of chemistry, properties and crystallisation, J. Eur. Ceram. Soc. 32, 1925-1932 (2012).

[117] F. Muñoz, A. Durán, L. Pascual, R. Marchand: Compositional and viscosity

influence on the nitrogen/oxygen substitution reactions in phosphate melts, Phys. Chem.

Glasses 46 [1], 39-45 (2005).

[118] R. Marchand, D. Agliz, L. Boukbir, A. Quemerais: Characterization of nitrogen

containing phosphate glasses by X-ray photoelectron spectroscopy, J. Non-Cryst. Solids

103, 35-44 (1988).

[119] B.C. Bunker, G.W. Arnold, M. Rajaram, D.E. Day: Corrosion of phosphorus

oxynitride glasses in water and Humid air, J. Am. Ceram. Soc. 70, 425-430 (1987).

[120] L. Pascual, A. Durán: Preparation and properties of nitride phosphate glasses,

Glastech. Ber. 64 [2], 43-48 (1991).

[121] M.R. Reidmeyer, D.E. Day: Phosphorus oxynitride glasses, J. Non-Cryst. Solids

181, 201-214 (1995).

[122] A. Le Sauze, R. Marchand: Chemically durable nitride phosphate glasses

resulting from nitrogen/oxygen substitution within PO4 tetrahedra, J. Non-Cryst. Solids

263&264, 285-292 (2000).

[123] G.L. Paraschiv, F. Muñoz, L.R. Jensen, Y. Yue, M. Smedskjaer: Impact of

nitridation of metaphosphate glasses on liquid fragility, J. Non-Cryst. Solids 441, 22-28

(2016).

[124] M.R. Reidmeyer, D.E. Day: Preparation and properties of nitrogen-doped

phosphate glasses, J. Am. Ceram. Soc. 68 [8] C-188-C-190 (1985).

104

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [125] R.K. Brow, M.R. Reidmeyer, D.E. Day: Oxygen bonding in nitrided sodium- and

lithium-metaphosphate glasses, J. Non-Cryst. Solids 99, 178-189 (1988).

[126] A. Le Sauze, L. Montagne, G. Palavit, F. Fayon, R. Marchand : X-ray

photoelectron spectroscopy and nuclear magnetic resonance structural study of

phosphorus oxynitride glasses ‘LiNaPON’, J. Non-Cryst. Solids 263&264, 139-145

(2000).

[127] F. Muñoz, L. Pascual, A. Durán, J. Rocherullé, R. Marchand : Alkali and alkali-

lead oxynitride phosphate glasses : a comparative structural study by NMR and XPS,

C.R. Chimie 5, 731-738 (2002).

[128] F. Muñoz, L. Pascual, A. Durán, L. Montagne, G. Palavit, R. Berjoan, R.

Marchand : Structural of phosphorus oxynitride glasses LiNaPbPON by nuclear

magnetic resonance and X-ray photoelectron spectroscopy, J. Non-Cryst. Solids 324,

142-149 (2003).

[129] F. Muñoz, L. Pascual, A. Durán, R. Berjoan, R. Marchand : Validation of the

mechanism of nitrogen/oxygen substitution in Li-Na-Pb-P-O-N oxynitride phosphate

glasses, J. Non-Cryst. Solids 352, 3947-3951 (2006).

[130] F. Muñoz: Kinetic analysis of the substitution of nitrogen for oxygen in phosphate

glasses, Phys. Chem. Glasses: Eur. J. Glass Sci. Technol. B 52 [4], 181-186 (2011).

[131] F. Muñoz, L. Delevoye, L. Montagne, T. Charpentier: New insights into the

structure of oxynitride NaPON phosphate glasses by 17-oxygen NMR, J. Non-Cryst.

Solids 363, 134-139 (2013).

[132] F. Muñoz, A. Durán, L. Pascual: Synthesis and properties of nitrided phosphate

glasses in the system R2O-R’O-PbO-P2O5 (R,R’=Li,Na), Phys. Chem. Glasses 43C,

113-118 (2002).

105

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [133] Q. Riguidel, F. Muñoz: Effect of nitridation on the aqueous dissolution of Na2O-

K2O-CaO-P2O5 metaphosphate glasses, Acta Biomaterialia 7, 2631-2636 (2011).

[134] B. Wang, B.S. Kwak, B.C. Sales, J.B. Bates: Ionic conductivities and structure of

lithium phosphorus oxynitride glasses, J. Non-Cryst. Solids, 183, 297-306 (1995).

[135] R.K. Brow, D.R. Tallant: Structural design of sealing glasses, J. Non-Cryst.

Solids 222, 396-406 (1997).

[136] J. Rocherullé, J. Massera, H. Oudadesse, L. Calvet, J. Trolès, X.H. Zhang : Heat

capacities of crystalline and glassy lithium metaphosphate up to the transition region, J.

Therm. Anal. Calorim. 123, 401-407 (2016).

[137] S. Inaba, S. Oda, K. Morinaga : Heat capacity of oxide glasses measured by AC

calorimetry, J. Non-Cryst. Solids 306, 42-49 (2002).

[138] J.J. Hudgens, R.K. Brow: Raman spectroscopy study of the structure of lithium

and sodium ultraphosphate glasses, J. Non-Cryst. Solids 223, 21-31 (1998).

[139] T.M. Alam, S. Conzone, R.K. Brow, T.J. Boyle: 6Li, 7Li nuclear magnetic

resonance investigation of lithium coordination in binary phosphate glasses, J. Non-

Cryst. Solids 258, 140-154 (1999).

[140] R.G. Frieser: A review of solder glasses, Electrocomponent Sci. Technol. 2, 163-

199 (1975).

[141] I.W. Donald: Preparation, properties and chemistry of glass and glass-ceramic-to-

metal seals and coatings, J. Mater. Sci. 28, 2841-2886 (1993)

[142] J. A. Wilder, J. T. Healey, B. C. Bunker, in "Advances in Ceramics" Vol. 4,

edited by J. H. Simmons (American Ceramic Society, Columbus OH) 313-326 (1982).

[143] T.H. Wang, P. F. James, in "Proceedings of the 2nd International Conference on

New Materials and their Applications", University of Warwick, UK, 10-12 April 1990,

106

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 Institute of Physics Conference Series no. 111, edited by D. Holland (IOP, Bristol) 401-

410 (1990).

[144] I. W. Donald, P. M. Mallinson, B. L. Metcalfe, L. A. Gerrard, J. A. Fernie: Recent

developments in the preparation, characterization and applications of glass and glass-

ceramic-to-metal seals and coatings, J. Mater. Sci. 46, 1975-2000 (2011).

[145] A.E. Marino, S.R. Arrasmith, L.L. Gregg, S.D. Jacobs, G. Chen, Y. Duc: Durable

phosphate glasses with lower transition temperatures, J. Non-Cryst. Solids 289, 37-41

(2001).

[146] D. Ehrt: Phosphate and Fluoride-phosphate Optical Glasses –Properties, Structure

and Applications, Phys. Chem. Glasses: Eur. J. Glass Sci. Technol. B 56 [6], 217-234

(2015).

[147] E.T.Y. Lee, E.R.M. Taylor: Thermo-optic coefficients of potassium alumino-

metaphosphate glasses, J. Non-Cryst.Solids 65 1187-1192 (2004).

[148] A.K. Varshneya: Fundamentals of Inorganic Glasses (Academic Press, INC.,

London 1994).

[149] P.R. Ehrmann, K. Carlson, J.H. Campbell, C.A. Click, R.K. Brow: Neodymium

fluorescence quenching by hydroxyl groups in phosphate laser glasses, J. Non-Cryst.

Solids 349, 105-114 (2004).

[150] J.S. Hayden, M.K. Aston, S.A. Payne, M.L. Elder, J.H. Campbell: Laser and

thermo-physical properties of Nd-doped phosphate glasses, Proceedings of SPIE 1761,

162-173 (1992).

[151] G. Ofelt: Intensities of Crystal Spectra of Rare-Earth Ions, J. Chem. Phys. 37,

511-520 (1962).

[152] B. Judd: Optical absorption intensities of rare-earth ions, Phys. Rev. 127, 750-761

(1962).

107

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [153] W.F. Krupke: Induced-emission cross sections in neodymium laser glasses, IEEE

Journal of Quantum Electronics 10, 450-457 (1974).

[154] C. Thorsness, T.I. Suratwala, R.A. Steele, J.H. Campbell, J.S. Hayden, S.

Pucilowski, K. Suzuki: Dehydroxylation of phosphate laser glass, International

Symposium on Optical Science and Technology, International Society for Optics and

Photonics, 175-194 (2000).

[155] P. Ehrmann, J. Campbell, T. Suratwala, J. Hayden, D. Krashkevich, K. Takeuchi:

Optical loss and Nd3+ non-radiative relaxation by Cu, Fe and several rare-earth

impurities in phosphate laser glasses, J. Non-Cryst. Solids 263, 251-262 (2000).

[156] V. Arbuzov, Y.K. Fyodorov, S. Kramarev, S. Lunter, S. Nikitina, A. Pozharskii, A.

Shashkin, A. Semyonov, V. Ter-Nersesyants, A. Charukhchev: Neodymium phosphate

glasses for the active elements of a 128 channel laser facility, Glass Technology 46, 67-

70 (2005).

[157] L. Hu, S. chen, J. Tang, B. Wang, T. Meng, W. Chen, L. Wen, J. Hu, S. Li, Y.

Xu, Y. Jiang, J. Zhang, Z. Jiang: Large aperture N31 neodymium phosphate laser glass

for use in a high power laser facility, High Power Laser Science and Engineering 2, 1-6

(2014).

[158] E.A. Abou Neel, W. Chrzanowski, D.M. Pickup, L.A. O'Dell, N.J. Mordan, R.J.

Newport, M.E. Smith, J.C. Knowles: Structure and properties of strontium-doped

phosphate-based glasses, Journal of The Royal Society Interface 6, 435-446 (2009).

[159] E.A. Abou Neel, I. Ahmed, J.J. Blaker, A. Bismarck, A.R. Boccaccini, M.P.

Lewis, S.N. Nazhat, J.C. Knowles: Effect of iron on the surface, degradation and ion

release properties of phosphate-based glass fibres, Acta Biomaterialia 1, 553-563

(2005).

108

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [160] E.A. Abou Neel, W. Chrzanowski, J.C. Knowles: Effect of increasing titanium

dioxide content on bulk and surface properties of phosphate-based glasses, Acta

Biomaterialia 4, 523-534 (2008).

[161] M. Navarro, M.P. Ginebra, J.A. Planell: Cellular response to calcium phosphate

glasses with controlled solubility, J Biomed Mater Res A 67, 1009-15 (2003).

[162] D. Furniss, A.B. Seddon: Towards monomode proportioned fibreoptic preforms

by extrusion, Journal of Non-Crystalline Solids 256–257, 232-236 (1999).

[163] F.T. Wallenberger, N.E. Weston: Glass Fibers from High and Low Viscosity

Melts, Mat Res Soc Symposium Proc. 702, 165-172 (2002).

[164] J. Choueka, J.L. Charvet, H. Alexander, Y.O. Oh, G. Joseph, N.C. Blumenthal,

W.C. LaCourse: Effect of annealing temperature on the degradation of reinforcing

fibers for absorbable implants, J Biomed Mater Res. 29, 1309-1315 (1995).

[165] I. Ahmed, C.A. Collins, M.P. Lewis, I. Olsen, J.C. Knowles: Processing,

characterisation and biocompatibility of iron-phosphate glass fibres for tissue

engineering, Biomaterials 25, 3223-3232 (2004).

[166] I. Ahmed, S.S. Shaharuddin, N. Sharmin, D. Furniss, C. Rudd, Core/Clad

Phosphate Glass Fibres Containing Iron and/or Titanium, Biomedical glasses 1, 20-30

(2015).

[167] F. Ungaro, R. d'Emmanuele di Villa Bianca, C. Giovino, A. Miro, R. Sorrentino,

F. Quaglia, M.I. La Rotonda: Insulin-loaded PLGA/cyclodextrin large porous particles

with improved aerosolization properties: In vivo deposition and hypoglycaemic activity

after delivery to rat lungs, Journal of Controlled Release 135, 25-34 (2009).

[168] T. Rouxel: Elastic properties of glasses: a multiscale approach, Comptes Rendus

Mécanique 334, 743-753 (2006).

109

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [169] M. Goldstein, T.H. Davies: Glass Fibers with Oriented Chain Molecules, Journal

of the American Ceramic Society 38, 223-226 (1955).

[170] S. Inaba, H. Hosono, S. Ito: Entropic shrinkage of an oxide glass, Nat Mater. 14,

312-317 (2015).

[171] I. Ahmed, P.S. Cronin, E.A. Abou Neel, A.J. Parsons, J.C. Knowles, C.D. Rudd:

Retention of mechanical properties and cytocompatibility of a phosphate-based glass

fiber/polylactic acid composite, J Biomed Mater Res B Appl Biomater. 89, 18-27

(2009).

[172] I. Ahmed, I. Jones, A. Parsons, J. Bernard, J. Farmer, C. Scotchford, G. Walker,

C. Rudd: Composites for bone repair: phosphate glass fibre reinforced PLA with

varying fibre architecture, Journal of Materials Science: Materials in Medicine 22,

1825-1834 (2011).

[173] A.Z. Kharazi, M.H. Fathi, F. Bahmany: Design of a textile composite bone plate

using 3D-finite element method, Materials & Design 31, 1468-1474 (2010).

[174] S.L. Evans, P.J. Gregson: Composite technology in load-bearing orthopaedic

implants, Biomaterials 19, 1329-1342 (1998).

[175] N. Sharmin, A.J. Parsons, C.D. Rudd, I. Ahmed: Effect of boron oxide addition

on fibre drawing, mechanical properties and dissolution behaviour of phosphate-based

glass fibres with fixed 40, 45 and 50 mol% P(2)O(5), Journal of Biomaterials

Applications 29, 639-653 (2014).

[176] R.M. Felfel, I. Ahmed, A.J. Parsons, P. Haque, G.S. Walker, C.D. Rudd:

Investigation of Crystallinity, Molecular Weight Change, and Mechanical Properties of

PLA/PBG Bioresorbable Composites as Bone Fracture Fixation Plates, Journal of

Biomaterials Applications 26, 765-789 (2012).

110

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [177] R. Shah, A.C.M. Sinanan, J.C. Knowles, N.P. Hunt, M.P. Lewis: Craniofacial

muscle engineering using a 3-dimensional phosphate glass fibre construct, Biomaterials

26, 1497-1505 (2005).

[178] E.A. Abou Neel, I. Ahmed, J. Pratten, S.N. Nazhat, J.C. Knowles:

Characterisation of antibacterial copper releasing degradable phosphate glass fibres,

Biomaterials 26, 2247-2254 (2005).

[179] M. Yamaguchi, H. Oishi, Y. Suketa: Stimulatory effect of zinc on bone formation

in tissue culture, Biochemical Pharmacology 36, 4007-4012 (1987).

[180] C. Gérard, L.-J. Bordeleau, J. Barralet, C.J. Doillon: The stimulation of

angiogenesis and collagen deposition by copper, Biomaterials 31, 824-831 (2010).

[181] C. Vitale-Brovarone, G. Novajra, J. Lousteau, D. Milanese, S. Raimondo, M.

Fornaro: Phosphate glass fibres and their role in neuronal polarization and axonal

growth direction, Acta Biomaterialia 8, 1125-1136 (2012).

[182] S.N. Nazhat, E.A. Abou Neel, A. Kidane, I. Ahmed, C. Hope, M. Kershaw, P.D.

Lee, E. Stride, N. Saffari, J.C. Knowles, R.A. Brown: Controlled Microchannelling in

Dense Collagen Scaffolds by Soluble Phosphate Glass Fibers, Biomacromolecules 8,

543-551 (2006).

[183] N.-Y. Joo, J.C. Knowles, G.-S. Lee, J.-W. Kim, H.-W. Kim, Y.-J. Son, J.K. Hyun:

Effects of phosphate glass fiber–collagen scaffolds on functional recovery of completely

transected rat spinal cords, Acta Biomaterialia 8, 1802-1812 (2012).

[184] R. Kayacan: The effect of staining on the monotonic tensile mechanical properties

of human cortical bone, Journal of Anatomy 211, 654-661 (2007).

[185] F.G. Evans: Mechanical properties and histology of cortical bone from younger

and older men, The Anatomical Record 185, 1-11 (1976).

111

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [186] R. Felfel, I. Ahmed, A. Parsons, L. Harper, C. Rudd: Initial mechanical properties

of phosphate-glass fibre-reinforced rods for use as resorbable intramedullary nails,

Journal of Materials Science 47, 4884-4894 (2012).

[187] R.M. Felfel, I. Ahmed, A.J. Parsons, C.D. Rudd: Bioresorbable screws reinforced

with phosphate glass fibre: Manufacturing and mechanical property characterisation,

Journal of the Mechanical Behavior of Biomedical Materials 17, 76-88 (2013).

[188] C.A. Scotchford, M. Shataheri, P.S. Chen, M. Evans, A.J. Parsons, G.A.

Aitchison, C. Efeoglu, J.L. Burke, A. Vikram, S.E. Fisher, C.D. Rudd: Repair of

calvarial defects in rats by prefabricated, degradable, long fibre composite implants, J.

Biomed. Mater. Res. A. 96, 230-238 (2010).

[189] A. Alani, J.C. Knowles, W. Chrzanowski, Y.L. Ng, K. Gulabivala: Ion release

characteristics, precipitate formation and sealing ability of a phosphate glass-

polycaprolactone-based composite for use as a root canal obturation material, Dent

Mater. 25, 400-410 (2009).

[190] T. Ohtomo, F. Mizuno, A. Hayashi, K. Tadanaga, M. Tatsumisago: Electrical and

electrochemical properties of Li2S-P2S5-P2O5 glass-ceramic electrolytes, J. Power

Sources 146, 715-718 (2005).

[191] S. Chenu, R. Lebullenger, P. Bérnard-Rocherullé, G. Calvez, O. Guillou, J.

Rocherullé, A. Kidari, M.J. Pomeroy, S. Hampshire: Glass reactive sintering as an

alternative route for the synthesis of NZP glass-ceramics, J. Mat. Sci. 47, 486-492

(2012).

[192] G. Delaizir, V. Seznec, P. Rozier, C. Surcin, P. Salles, M. Dollé : Electrochemical

performances of vitreous materials in the system Li2O-V2O5-P2O5 as electrode for

lithium batteries, Solid State Ionics 237, 22-27 (2013).

112

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [193] S.W. Martin, C.A. Angell: Dc and Ac conductivity in wide composition range

Li2O-P2O5 glasses, J. Non-Cryst. Solids 83, 185-207 (1986).

[194] H. Takahashi, H. Nakanii, T. Sakuma, Y. Onoda: Sodium ion motion in NaI-

AgPO3 glasses, Solid State Ionics 179, 2137-2141 (2008).

[195] A.C.M. Rodrigues, M.L.F. Nascimento, C.B. Bragatto, J.-L. Souquet: Charge

carrier mobility and concentration as a function of composition in AgPO3-AgI glasses,

J. Chem. Phys. 135, 234504 (2011).

[196] O.L. Anderson, D.A. Stuart: Calculation of activation energy of ionic

conductivity in silica glasses by classical methods, J. Am. Ceram. Soc. 37 [12], 573-580

(1954).

[197] M.D. Ingram, C.T. Moynihan, A.V. Lesikar: Ionic conductivity and the weak

electrolyte theory of glass, J. Non-Cryst. Solids, 38&39, 371-376 (1980).

[198] B. Santic, A. Mogus-Milankovic, D.E. Day: The dc electrical conductivity of iron

phosphate glasses, J. Non-Cryst. Solids 296, 65-73 (2001).

[199] M. Wasiucionek, J.E. Garbarczyk, P. Kurek, J. Jakubowski: Electrical properties

of glasses of the Na2O-V2O5-P2O5 system, Solid State Ionics 70&71, 346-349 (1994).

[200] J.E. Garbarczyk, P. Jozwiak, M. Wasiucionek, J.L. Nowinski: Effect of

nanocrystallization on the electronic conductivity of vanadate-phosphate glasses, Solid

State Ionics 177, 2585-2588 (2006).

[201] A.K. Kercher, J.O. Ramey, K.J. Carroll, J.O. Kiggans, N.J. Dudney, R.A.

Meisner, L.A. Boatner, G.M. Weith: Mixed polyanion glass cathodes: iron phosphate

vanadate glasses, J. Electrochem. Soc. 161 [14], A2210-A2215 (2014).

[202] A.K. Kercher, J.A. Kolopus, K.J. Carroll, R.R. Unocic, S. Kirklin, C. Wolverton,

S.L. Stooksbury, L.A. Boatner, N.J. Dudney: Mixed polyanion glass cathodes: glass-

state conversion reactions, J. Electrochem. Soc. 163 [2], A131-A137 (2016).

113

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [203] J.C. Bazan, J.A. Duffy, M.D. Ingram, M.R. Mallace: Conductivity anomalies in

tungstate-phosphate glasses: evidence for an ion-polaron interaction?, Solid State Ionics

86-88, 497-501 (1996).

[204] I. Oliva, A. Masuno, H. Inoue, H. Tawarayama, H. Kawazoe: Mixed conduction

in alkali niobium tungsten phosphate glasses, Solid State Ionics, 206, 405-409 (2012).

[205] Y. Abe, H. Shimakawa: Protonic conduction in alkaline earth metaphosphate

glasses containing, J. Non-Cryst. Solids 51, 357-365 (1982).

[206] H. Sumi, Y. Nakano, Y. Fujishiro, T. Kasuga: Proton conduction of MO-P2O5

glasses (M=Zn,Ba) containing a large amount of water, Solid State Sciences 45, 5-8

(2015).

[207] N.J. Dudney: Addition of a thin-film inorganic electrolyte (Lipon) as a protective

film in lithium batteries with a liquid electrolyte, J. Power Sources 89, 176-179 (2000).

[208] J.B. Bates, N.J. Dudney, B. Neudecker, A. Ueda, C.D. Evans: Thin-film lithium

and lithium-ion batteries, Solid State Ionics 135, 33-45 (2000).

[209] S. Jacke, J. Song, L. Dimesso, J. Brötz, D. Becker, W. Jaegermann: Temperature

dependent phosphorus oxynitride growth for all-solid-state batteries, J. Power Sources

196, 6911-6914 (2011).

[210] F. Muñoz: Comments on the structure of LiPON thin-film electrolytes, J. Power

Sources 198, 432-433 (2012).

[211] F. Muñoz, A. Durán, L. Pascual, L. Montagne, B. Revel, A.C.M. Rodrigues:

Increased electrical conductivity of LiPON glasses produced by ammonolysis, Solid

State Ionics 179, 574-579 (2008).

[212] N. Mascaraque, J.L.G. Fierro, A. Durán, F. Muñoz: An interpretation for the

increase of ionic conductivity by nitrogen incorporation in LiPON oxynitride glasses,

Solid State Ionics 233, 73-79 (2013).

114

Muñoz, F., Rocherullé, J., Ahmed, I., Hu, L. (2019). Phosphate Glasses. En Musgraves, J. D., Hu, J., Calvez, L.. Springer Handbook of Glass (Springer Handbooks, pp. 553-594). Cham: Springer International Publishing. 10.1007/978-3-319-93728-1_16 [213] Y. Shimonishi, T. Zhang, N. Imanishi, D. Im, D.J. Lee, A. Hirano, Y. Takeda, O.

Yamamoto, N. Sammes: A study on lithium/air secondary batteries-Stability of the

NASICON-type lithium ion conducting solid electrolyte in alkaline aqueous solutions,

J. Power Sources 196, 5128-5132 (2011).

[214] S. Hasegawa, N. Imanishi, T. Zhang, J. Xie, A. Hirano, Y. Takeda, O. Yamamoto:

Study on lithium/air secondary batteries-Stability of NASICON-type lithium ion

conducting glass-ceramics with water, J. Power Sources 189, 371-377 (2009).

[215] X. Yu, D.E. Day, G.J. Long, R.K. Brow: Properties and structure of sodium-iron

phosphate glasses, J. Non-Cryst. Solids 215, 21-31 (1997).

[216] D.E. Day, Z. Wu, C.S. Ray, P. Hrma: Chemically durable iron phosphate glass

wasteforms, J. Non-Cryst. Solids 241, 1-12 (1998).

[217] M.I. Ojovan, W.E. Lee: An introduction to nuclear waste immobilization

(Elsevier Science Publishers B.V., Amsterdam 2005).

[218] C.W. Kim, C.S. Ray, D. Zhu, D.E. Day, D. Gombert, A. Aloy, A. Mogus-

Milankovi, M. Karabulut: Chemically durable iron phosphate glasses for vitrifying

sodium bearing waste (SBW) using conventional and cold crucible induction melting

(CCIM) techniques, Journal of Nuclear Materials 322, 152-164 (2003).

115